Author Topic: Reducing sequence of returns risk with margin  (Read 5260 times)

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Reducing sequence of returns risk with margin
« on: November 25, 2021, 08:22:38 AM »
I was thinking about how billionaires can live on just margin, without selling their stocks, when I realized there might be something valuable for us non-billionaires.  If you retire, and the market crashes - take out a margin loan for living expenses.

Maybe someone retires with $1.5 million, and plans to follow the 4% rule and withdraw $60k/year.  But after a crash of 50%, they have $750k in assets and still need $60k/year, which is now an 8% withdrawal rate.  But they could wait out the crash, living on money withdrawn from a margin account.

Using very rough numbers, the crash + recovery might take 3 years.  Margin loans might be more expensive, so call it $7k interest and $180k in living expenses.  This person pays off their $187k margin loan once their stocks are back to $1.5 million.  They sell $187k of stock in one year and pay off the margin loan.  They avoid sequence of returns risk using margin.

So maybe that's one way for people to avoid sequence of returns risk - when a crash occurs, withdraw living expenses from a margin account, and pay it back when stocks recover.

tooqk4u22

  • Magnum Stache
  • ******
  • Posts: 2846

Rob_bob

  • Bristles
  • ***
  • Posts: 406
  • Location: Oregon
Re: Reducing sequence of returns risk with margin
« Reply #2 on: November 25, 2021, 11:01:27 AM »
I wouldn't want to use margin in a falling market, you have interest to pay and an increasing chance of a margin call.

daverobev

  • Magnum Stache
  • ******
  • Posts: 3962
  • Location: France
Re: Reducing sequence of returns risk with margin
« Reply #3 on: November 25, 2021, 02:58:19 PM »
I wouldn't want to use margin in a falling market, you have interest to pay and an increasing chance of a margin call.

If you start with 0 margin, the market falls by 50% and you borrow 12% of the original balance you're only at 24% - you're not going to get called.

Conversely if the downturn lasts a decade... you're maybe just delaying the inevitable.

vand

  • Handlebar Stache
  • *****
  • Posts: 2339
  • Location: UK
Re: Reducing sequence of returns risk with margin
« Reply #4 on: November 26, 2021, 04:19:56 AM »
The whole problem with this line of thought is that you are assuming that everything else in the world stays the same while stocks are plummetting. That doesn't happen in the real world. People run for cover.

Do you think that in the teeth of the next panic when it seems our way of life is about to change, that someone is going to lend you money at the 1% or 2% rates that are available today? Hell no, borrowing costs go up, up up. If you go broke, they lose their money. So people will want 10% or 15%, or more to risk their capital.


The simplest thing you can do to protect yourself from sequence risk is to add a slice of gold. In the 2 post-war decade-ish long periods where stock/bond portfolios have struggled, gold performed as the lifeboat that would have kept the portfolio above water and improved its survivability on both occasions. That's a pretty good track record as a hedge against struggling paper assets.

According to cfiresim a 10% allocation to gold improves the survival rate of the portfolio (using standard parameters) from 95.87% to 99.17% for both 60/40 -> 60/30/10 and 75/25 -> 75/15/10.

It's about as big of no brainer as you can get in this discussion, I feel.
« Last Edit: November 26, 2021, 05:38:29 AM by vand »

daverobev

  • Magnum Stache
  • ******
  • Posts: 3962
  • Location: France
Re: Reducing sequence of returns risk with margin
« Reply #5 on: November 26, 2021, 05:44:35 AM »
Paper gold or real gold? If real, in the UK yeah you can do it with sovereigns and not worry about CGT which is pretty nice. But where are you going to store your £30k-100k of gold sovereigns?

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #6 on: November 26, 2021, 06:56:15 AM »
ERN had a recent post about this

https://earlyretirementnow.com/2021/11/16/leverage-in-retirement-swr-series-part-49/
Interesting, thanks.  It looks like the website is a one person operation, so I assume that article was written by the CFA who set it up.  I'd say a CFA has more credibility than my estimate.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #7 on: November 26, 2021, 06:59:15 AM »
I wouldn't want to use margin in a falling market, you have interest to pay and an increasing chance of a margin call.
Can you give a specific scenario?

Margin calls occur when the borrowing gets close to your amount of assets.  But if your spending reaches close to your level of assets, you'd have wiped out your retirement anyways.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #8 on: November 26, 2021, 07:02:11 AM »
Do you think that in the teeth of the next panic when it seems our way of life is about to change, that someone is going to lend you money at the 1% or 2% rates that are available today? Hell no, borrowing costs go up, up up. If you go broke, they lose their money. So people will want 10% or 15%, or more to risk their capital.
Do you have evidence Interactive Brokers charged 15% margin loans at any time?

In my example, I assumed 1% higher rates, but the theory still works at 10%, because stocks are losing faster than 10% in a crash.


The simplest thing you can do to protect yourself from sequence risk is to add a slice of gold. In the 2 post-war decade-ish long periods where stock/bond portfolios have struggled, gold performed as the lifeboat that would have kept the portfolio above water and improved its survivability on both occasions. That's a pretty good track record as a hedge against struggling paper assets.
This is a discussion about margin to avoid sequence of returns risk, not gold investing.  If you're really interested in gold, you should start a thread discussing it, and tracking how it helps sequence of returns risk.

index

  • Pencil Stache
  • ****
  • Posts: 656
Re: Reducing sequence of returns risk with margin
« Reply #9 on: November 26, 2021, 09:57:50 AM »
This would work with today's margin requirements and interest rate. The issue is will this work in a future scenario? Interactive Brokers and others can change margin requirements and interest rates as necessary to protect their core business. In the depth of a panic, who knows if and at what terms you can withdraw cash on margin.

Many of your posts seem to have the theme of finding a shortcut to raise investment returns - Crypto, 90% passive 10% active,  factor investing. There is not such thing as a free lunch. Risk can be diversified and you can get lucky by picking the next best sector, factor, or vehicle (crypto), but you equally run the chance of pegging that tilt at its high and underperforming.

Please buy crypto, invest 10% active, stay 100% invested in risk assets and plan on using margin when there is a 50% drawdown. Understand the risk you are taking versus a boring bond/equity mix. If the chance to outperform the boring portfolio is worth it, and you can stomach the volatility and unknowns then go for it. This is no different than arguing for a 5-6-7-8% withdraw rate in order to retire sooner or a yolo on bitcoin. There is a formula that has worked, you could get lucky and your idea works or you could end up toiling away at your day job longer than the sucker following the boring formula. 

mistymoney

  • Handlebar Stache
  • *****
  • Posts: 2431
Re: Reducing sequence of returns risk with margin
« Reply #10 on: November 26, 2021, 10:14:17 AM »
I wouldn't want to use margin in a falling market, you have interest to pay and an increasing chance of a margin call.

If you start with 0 margin, the market falls by 50% and you borrow 12% of the original balance you're only at 24% - you're not going to get called.

Conversely if the downturn lasts a decade... you're maybe just delaying the inevitable.

or intensifying it? such as the market drops 30% and stays low for a few years and then drops 15% again for another few years, and then you also have all the margin interest.

Aside from the eventual outcome, would the heartburn be worth it?

I think there are many ways margin could reduce the SoRR, but also many scenarios in which it could increase the risk.

Everything comes back to risk/reward, doesn't it? the additional risk would likely be rewarded (the average 3 year downturn) but it also upens you up to greater risk if the downturn is not average.

whywork

  • Stubble
  • **
  • Posts: 207
Re: Reducing sequence of returns risk with margin
« Reply #11 on: November 26, 2021, 04:48:28 PM »
I think this is a good approach. If market is 50% down, there is no point selling your investments.

I wonder what should be the general approach to using margin when the market is down say 10%, 20%, 30% in an year? Let's say we withdraw yearly once and we see for the previous year we are down by say x%. How much should we withdraw from margin that year?

If we use the same math then, if market is down 50%, we make all our annual withdrawal from margin. If market is down 25%, we withdraw half of our annual expenses from margin and half from selling investments. Similarly 12% market down we withdraw 25% of annual expenses from margin



ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 6733
  • Location: A poor and backward Southern state known as minimum wage country
Re: Reducing sequence of returns risk with margin
« Reply #12 on: November 26, 2021, 08:48:09 PM »
ERN had a recent post about this

https://earlyretirementnow.com/2021/11/16/leverage-in-retirement-swr-series-part-49/
Interesting, thanks.  It looks like the website is a one person operation, so I assume that article was written by the CFA who set it up.  I'd say a CFA has more credibility than my estimate.
ERN was writing about carrying a margin loan from the start of one's retirement. However, the OP by @MustacheAndaHalf proposed a conditional or algorithmic approach - only going into margin AFTER a huge correction or bear market.

I wrote ERN asking if he could retool his spreadsheet to allow for algorithmic approaches. My question was about retiring on a conservative AA and then switching to an aggressive AA when a certain trigger point was reached, such as a xx% drop from the top. He didn't want to do it and referred me to his forum - to my great disappointment. Such an alteration would also enable the answering of @MustacheAndaHalf 's question.

If any spreadsheet wizards have a few hours to spare this weekend, ERN's SWR spreadsheet is here:

https://docs.google.com/spreadsheets/d/1QGrMm6XSGWBVLI8I_DOAeJV5whoCnSdmaR8toQB2Jz8/edit

The task is:
1) Set up a trigger system where a cell flips to "yes" if the stock index's value drops by the specified amount below it's highest value up until that point.
2) Set up editable before-event AA and after-event AA fields. Set up an editable field to enter the threshold at which we switch AA's.
3) Replace the SWR calculation from that point on with the new AA, starting value, and remaining timeframe.

Something tells me either of our ideas could "rescue" the 4% rule.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #13 on: November 27, 2021, 05:20:00 AM »
Many of your posts seem to have the theme of finding a shortcut to raise investment returns - Crypto, 90% passive 10% active,  factor investing. There is not such thing as a free lunch. Risk can be diversified and you can get lucky by picking the next best sector, factor, or vehicle (crypto), but you equally run the chance of pegging that tilt at its high and underperforming.
The forum rules say to attack an idea, not the person.  This has nothing to do with the current thread - you're simply attacking me.  I've reported you for doing so - please re-read the forum rules.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #14 on: November 27, 2021, 05:59:03 AM »
I wouldn't want to use margin in a falling market, you have interest to pay and an increasing chance of a margin call.

If you start with 0 margin, the market falls by 50% and you borrow 12% of the original balance you're only at 24% - you're not going to get called.

Conversely if the downturn lasts a decade... you're maybe just delaying the inevitable.
or intensifying it? such as the market drops 30% and stays low for a few years and then drops 15% again for another few years, and then you also have all the margin interest.

Aside from the eventual outcome, would the heartburn be worth it?

I think there are many ways margin could reduce the SoRR, but also many scenarios in which it could increase the risk.

Everything comes back to risk/reward, doesn't it? the additional risk would likely be rewarded (the average 3 year downturn) but it also upens you up to greater risk if the downturn is not average.
I haven't seen IBKR charge excessive margin loan rates - they charge 1.58% now.  Their low rates is causing other brokers to lose business to IBKR, so places like Schwab and Fidelity are giving very low rates of their own if people ask.  There's no need for them to panic, either - your broker can sell your assets at any time they choose to pay down the margin loan.  They try not to be arbitrary from what I've seen, but they put that right in the margin agreement.

So let's say two people with $1M living on $40k/year try both approaches.  The market crashes 30%, so Mr Margin borrows $40k and has $700k of assets ($660k).  Mr Sell sells $40k of their $700k, and has $660k in assets (and no margin loan).  If the market recovers from here, Mr Margin winds up with $1M and a $40k loan + $800 interest (call it 2%), and sells $40.8k of $1M and has $959,200 in assets and a paid off margin loan.  If Mr Sell's remaining assets recover, they have $943k of assets left.

After one year, $959k vs $943k.  If this goes on for 3 years, then ends, there's $700k with $125k margin loan for Mr Margin ($575k NW), while Mr Sell keeps selling, and has $580k left.  Mr Margin is behind on interest, but has $700k invested versus $580k for Mr Sell.  So let's say it ends now, after 3 years, and the market recovers (goes up +42.857%).

Mr Margin recovers to $1M, and pays off the $125k margin loan, and has $875k in assets.  Mr Margin is out exact living expenses plus margin interest on those expenses.  But Mr Sell sees $580k recover to $828.6k, about $46k poorer.  A crash shrinks Mr Sell's portfolio, but not their rent and expenses.  Their $40k expenses are now a 5.7% withdrawal rate, which causes their portfolio to be permanently smaller.

The Great Depression started in 1929, wiped out 90% of stock value, and lasted for many years.  The government learned a very harsh lesson to never let that happen again.  That's why the Fed took an active role in both 2008 and 2020, lowering rates to 0% and buying assets to provide liquidity.

index

  • Pencil Stache
  • ****
  • Posts: 656
Re: Reducing sequence of returns risk with margin
« Reply #15 on: November 27, 2021, 09:13:48 PM »
Many of your posts seem to have the theme of finding a shortcut to raise investment returns - Crypto, 90% passive 10% active,  factor investing. There is not such thing as a free lunch. Risk can be diversified and you can get lucky by picking the next best sector, factor, or vehicle (crypto), but you equally run the chance of pegging that tilt at its high and underperforming.
The forum rules say to attack an idea, not the person.  This has nothing to do with the current thread - you're simply attacking me.  I've reported you for doing so - please re-read the forum rules.

How am I attacking you? I am just calling attention that many of your thread topics are essentially asking about adding risky assets that have done well the last few years in order to increase return.  Using margin in the event of a drawdown is market timing and adding margin debt against your stache increases risk.

The whole premise of this thread is - should I use leverage when the market has bottomed? The answer is absolutely. Picking the bottom is a lot harder - is it 30% - 50% -75% down?

Did you go on margin at the bottom in March 2020? If not, what makes you think you would pick the bottom during the next large drawdown? 

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #16 on: November 28, 2021, 12:50:49 AM »
Many of your posts seem to have the theme of finding a shortcut to raise investment returns - Crypto, 90% passive 10% active,  factor investing. There is not such thing as a free lunch. Risk can be diversified and you can get lucky by picking the next best sector, factor, or vehicle (crypto), but you equally run the chance of pegging that tilt at its high and underperforming.
The forum rules say to attack an idea, not the person.  This has nothing to do with the current thread - you're simply attacking me.  I've reported you for doing so - please re-read the forum rules.

How am I attacking you? I am just calling attention that many of your thread topics are essentially asking about adding risky assets that have done well the last few years in order to increase return.  Using margin in the event of a drawdown is market timing and adding margin debt against your stache increases risk.

The whole premise of this thread is - should I use leverage when the market has bottomed? The answer is absolutely. Picking the bottom is a lot harder - is it 30% - 50% -75% down?

Did you go on margin at the bottom in March 2020? If not, what makes you think you would pick the bottom during the next large drawdown? 
You said "many of my posts", which is entirely about me.  You bring up my posts from other threads, which have nothing to do with this thread.  You are trying to post about me, and not the topic.

"2. Attack an argument, not a person."
https://forum.mrmoneymustache.com/forum-information-faqs/forum-rules/

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #17 on: November 28, 2021, 01:12:23 AM »
Many of your posts seem to have the theme of finding a shortcut to raise investment returns - Crypto, 90% passive 10% active,  factor investing. There is not such thing as a free lunch. Risk can be diversified and you can get lucky by picking the next best sector, factor, or vehicle (crypto), but you equally run the chance of pegging that tilt at its high and underperforming.
The forum rules say to attack an idea, not the person.  This has nothing to do with the current thread - you're simply attacking me.  I've reported you for doing so - please re-read the forum rules.
Did you go on margin at the bottom in March 2020? If not, what makes you think you would pick the bottom during the next large drawdown?
I called the bottom in these forums, and went from 60% to 100% equities on Mar 20 (Friday).  The market bottomed the next trading day, Monday Mar 23.  Once I added margin to my account, I went on margin for the most profitable week in ~35 years, then ended the loan.  The market ignored Covid-19 that I studied, giving me an advantage over the market.

My signature line refers to my Roth IRA investments, which became larger than my total NW from 2019.  My Roth IRA, by itself, more than doubled my NW.  Margin is not allowed in IRA accounts.

Which means I don't need to use margin loans when a crash occurs.  I can ignore a crash, or go on margin.  It won't matter for me.  But for others, sequence of returns risk is a very real problem, and that is also when it's easiest to calculate how margin can be used for living expenses during a crash.

Note also that you're comparing two irrelevant things, because you're too busy focusing on me, instead of the topic.  The topic isn't going on margin to profit in a crash.  The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.

lutorm

  • Pencil Stache
  • ****
  • Posts: 831
  • Location: About the middle of Sweden
Re: Reducing sequence of returns risk with margin
« Reply #18 on: November 28, 2021, 11:40:07 AM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #19 on: November 28, 2021, 12:22:13 PM »
ERN was writing about carrying a margin loan from the start of one's retirement. However, the OP by @MustacheAndaHalf proposed a conditional or algorithmic approach - only going into margin AFTER a huge correction or bear market.

I wrote ERN asking if he could retool his spreadsheet to allow for algorithmic approaches. My question was about retiring on a conservative AA and then switching to an aggressive AA when a certain trigger point was reached, such as a xx% drop from the top.
Reading further, I see what you mean.  The article starts out stating something similar to what I mentioned here, but then goes off into borrowing multiples of the portfolio for decades.  Millions are borrowed off a $1 million portfolio, which is unlike my suggestion here.  They did mention the following crash:

From Jan 1973 - Sep 1974 was a -46% correction, with a recovery by Dec 1976.

With current margin rates, margin turns out better in this scenario.  But with 7% treasuries (10 yr) + 6% brokers like to add, 13% margin loan rates would have fared worse than selling assets.

That brings up another point: what is a crash?  I view it as the total percentage loss, and not 10% even if it happens in a few days.  I was thinking somewhere in the 20-25% range, which means I don't consider 1973 to be a crash.  Wikipedia's view:

"There is no numerically specific definition of a stock market crash but the term commonly applies to declines of over 10% in a stock market index over a period of several days."
https://en.wikipedia.org/wiki/Stock_market_crash

Lacking 1973 data, I don't see any mention of sudden percentage drops.  So the 18.2% loss in 1973 was probably not 10% over several days.  If you only go on margin after a 20-25% drop, that starts in 1974 and is 99% recovered in 1975.  So it's more like 2 years on margin, with far lower interest costs.

That's perhaps something I could explore more: what percentage loss is a crash, and so would trigger margin loans (only to pay for expenses).  And what is the breakeven rate on margin loans, where selling assets and the margin loan approach have the same outcome?

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #20 on: November 28, 2021, 12:37:22 PM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.
How can losses be amplified by an approach that never buys assets?

Your claims of a safe withdrawal rate by definition ignore "sequence of returns risk", which is the whole point of this thread.  A "safe" withdrawal rate can result in going broke if a significant crash occurs very early on.
https://www.kitces.com/blog/understanding-sequence-of-return-risk-safe-withdrawal-rates-bear-market-crashes-and-bad-decades/

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with a loan
« Reply #21 on: November 28, 2021, 12:45:52 PM »
I've confused people with the word "margin" which often means leverage.  So forget about "margin loans", and consider a cheap bank loan for 2%.  In my scenario, it is exactly the same.

The question is if letting assets stay invested through a recovery is better than selling assets in a crash.  A crash early in retirement is often called "sequence of returns risk", and can derail retirement plans which seemed safe before the crash.

Two different retired people with $100k and no loans each encounter a 20% or greater crash.  One of them sells assets after the crash to pay for living expenses.  The other takes out a bank loan at 2% to pay for living expenses, and does not buy or sell assets.

That's the scenario I'm proposing: if a cheap bank loan for living expenses is better than selling assets during a crash.  Once the recovery occurs, the bank loan is paid off.  One person has constant expenses, in exchange for paying interest on the loan.  The other sold assets that are gone once the recovery occurs.  Their losses are increased by selling stocks that were heavily discounted.
« Last Edit: November 28, 2021, 12:47:53 PM by MustacheAndaHalf »

lutorm

  • Pencil Stache
  • ****
  • Posts: 831
  • Location: About the middle of Sweden
Re: Reducing sequence of returns risk with margin
« Reply #22 on: November 28, 2021, 04:40:26 PM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.
How can losses be amplified by an approach that never buys assets?

Your claims of a safe withdrawal rate by definition ignore "sequence of returns risk", which is the whole point of this thread.  A "safe" withdrawal rate can result in going broke if a significant crash occurs very early on.
https://www.kitces.com/blog/understanding-sequence-of-return-risk-safe-withdrawal-rates-bear-market-crashes-and-bad-decades/
Your losses are amplified when the margin loan gets called and your assets are sold. Selling all the assets at a bad time amplifies the loss.

I'm not sure what you mean that a SWR ignores SORR. There's a 1000-post thread on here about exactly that topic. The definition of a safe withdrawal rate is the largest withdrawal rate that does not run out of money in any historical scenario. IF you use historical scenarios, you also have SORR.

I see in your other post that you're now saying get a bank loan instead of a margin loan. But no bank would lend you money for your expenses without collateral. (Or rather, that would be a consumer loan with 15% interest rate.) The only reason a margin loan can get such a good interest rate is because the brokerage knows they can just sell your assets and pay back the loan if they don't like how things are shaping up.

Of course a loan will be cheaper in the situation you postulate, where the market recovers and you pay it back. The valid question is whether you can distinguish, in advance, that situation from one where the market doesn't recover until you've borrowed enough that your loan gets called.

JJ-

  • Pencil Stache
  • ****
  • Posts: 886
Re: Reducing sequence of returns risk with margin
« Reply #23 on: November 28, 2021, 05:01:16 PM »
I've been thinking about this for a few days and I am not sure how well it would work in practice. I like the idea of it.

A couple of things that I keep going back to. First, I feel like those that would be interested in leveraging in margin would already be leveraged to an extent. I plan on using some, but only up to a point where a large crash would not trigger a margin call.

Assuming that crash came, that's when the time an unleveraged person would leverage margin per your question.

Now what does an already leveraged person do? My assumption here is that you still need the bonds component  likely from another account (Roth IRA for example) to either pay off margin to get away from the call threshold or start pumping into the taxable account to increase liquidity.

So it may prevent you from needing to tent so heavily early on, but still need to tent.

You've probably already said all this but figured I'd chime in.

Niceday

  • 5 O'Clock Shadow
  • *
  • Posts: 73
Re: Reducing sequence of returns risk with margin
« Reply #24 on: November 28, 2021, 10:11:59 PM »
Very dangerous idea. If you had used this idea during the internet bubble which was followed by the 9/11 event, and the mortgage crisis, you would have been blown out of the game.

daverobev

  • Magnum Stache
  • ******
  • Posts: 3962
  • Location: France
Re: Reducing sequence of returns risk with margin
« Reply #25 on: November 29, 2021, 12:46:15 AM »
https://old.reddit.com/r/FIREUK/comments/r4i2fl/whats_longest_amount_of_time_to_breakeven/

Quote
If I invest into the s&p500 at the peak Price and the price falls and never put anymore money in, what’s the longest amount of time historically has it taken to breakeven?

Quote
Adjusted for inflation, it's about 1971 to 1984.

Don't think you'd want to be carrying a loan for 13 years - with reinvested dividends.

vand

  • Handlebar Stache
  • *****
  • Posts: 2339
  • Location: UK
Re: Reducing sequence of returns risk with margin
« Reply #26 on: November 29, 2021, 03:06:03 AM »
Paper gold or real gold? If real, in the UK yeah you can do it with sovereigns and not worry about CGT which is pretty nice. But where are you going to store your £30k-100k of gold sovereigns?

Well if the rest of the stash is in paper assets then I don't see the real drawback of having your gold position in paper too. But it doesn't have to be entirely one or the other - you can some of each. Citing the difficulties of storing physical is a naff but convenient excuse. £100k of gold sovereigns is about the size of your fist - instead of asking where are you going to store it, a better question is "where can't you store it?"

daverobev

  • Magnum Stache
  • ******
  • Posts: 3962
  • Location: France
Re: Reducing sequence of returns risk with margin
« Reply #27 on: November 29, 2021, 03:44:57 AM »
Paper gold or real gold? If real, in the UK yeah you can do it with sovereigns and not worry about CGT which is pretty nice. But where are you going to store your £30k-100k of gold sovereigns?

Well if the rest of the stash is in paper assets then I don't see the real drawback of having your gold position in paper too. But it doesn't have to be entirely one or the other - you can some of each. Citing the difficulties of storing physical is a naff but convenient excuse. £100k of gold sovereigns is about the size of your fist - instead of asking where are you going to store it, a better question is "where can't you store it?"

Sure, there are two different things I suppose, total global collapse where paper assets are worthless, you can't access a safety deposit box, etc, vs just a normal market crash where stuff keeps functioning.

For physical, I meant - for insurance purposes, and that you're probably going to want a safe that is fireproof, intruder alarm, etc, etc. I suppose these days having £30k in your house is not completely insane, I would be quite scared of having £100k though. Not totally unreasonable to just have a couple of dozen coins hidden in a wall cavity or under the floor boards, but then you have the risk of moving and forgetting about them...!

vand

  • Handlebar Stache
  • *****
  • Posts: 2339
  • Location: UK
Re: Reducing sequence of returns risk with margin
« Reply #28 on: November 29, 2021, 05:25:33 AM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.

^ This.

lutorm & index actually have a clue about what is being suggested here.

There are no free lunches. You may get lucky, but that one time it doesn't go your way will get you killed.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #29 on: November 29, 2021, 06:44:03 AM »
Very dangerous idea. If you had used this idea during the internet bubble which was followed by the 9/11 event, and the mortgage crisis, you would have been blown out of the game.
All of which are false, because that is not what I proposed.  I'm starting to suspect people are reading the title, and posting - without reading what I described.

Call a crash a 20% drawdown.  From Nov 2007 to July 2008, the "margin loan" approach does nothing because the drawdown is less than 20%.  In Aug 2008, the crash reaches -22%, and the "margin loan for living expenses" kicks in.  In the 6 months from Sep 2008 to Feb 2009, the market bottoms out with a -42% loss.  You have 58% of your original assets, and a 2% margin loan from 6 months of living expenses.  You have 29x your margin loan - there's no margin call.

Now you wait, and from Sep 2008 to Dec 2010 the market has a return of +6%.  After 2 years and 3 months, you end the margin loan.  If your withdrawal rate was 4%, you spent 9% on margin plus interest, and now have 106% of your starting balance to pay off that loan.  You have more than 11x the margin loan - again, no margin call.

In "the mortgage crisis" this approach bottoms out at 11x assets to loan, so it is incorrect to claim it "would have been blown out of the game".  You made a false claim without understanding what I proposed.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #30 on: November 29, 2021, 06:46:06 AM »
https://old.reddit.com/r/FIREUK/comments/r4i2fl/whats_longest_amount_of_time_to_breakeven/

Quote
If I invest into the s&p500 at the peak Price and the price falls and never put anymore money in, what’s the longest amount of time historically has it taken to breakeven?

Quote
Adjusted for inflation, it's about 1971 to 1984.

Don't think you'd want to be carrying a loan for 13 years - with reinvested dividends.
You also don't understand.  At the peak, this approach does nothing.  There's no margin loan.  Only after a 20% crash has already occurred, this approach does not buy or sell assets - it simply uses margin loans for living expenses.  The margin loan is only for cash, and is repaid during the recovery.

When has the market been down more than -20% for 13 years?

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #31 on: November 29, 2021, 06:54:33 AM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.
^ This.

lutorm & index actually have a clue about what is being suggested here.

There are no free lunches. You may get lucky, but that one time it doesn't go your way will get you killed.
If this was a class assignment, your post would get an "F" grade for plagiarism.

There is not such thing as a free lunch.
...
you could get lucky and your idea works or you could end up toiling away at your day job longer than the sucker following the boring formula.

daverobev

  • Magnum Stache
  • ******
  • Posts: 3962
  • Location: France
Re: Reducing sequence of returns risk with margin
« Reply #32 on: November 29, 2021, 06:57:43 AM »
https://old.reddit.com/r/FIREUK/comments/r4i2fl/whats_longest_amount_of_time_to_breakeven/

Quote
If I invest into the s&p500 at the peak Price and the price falls and never put anymore money in, what’s the longest amount of time historically has it taken to breakeven?

Quote
Adjusted for inflation, it's about 1971 to 1984.

Don't think you'd want to be carrying a loan for 13 years - with reinvested dividends.
You also don't understand.  At the peak, this approach does nothing.  There's no margin loan.  Only after a 20% crash has already occurred, this approach does not buy or sell assets - it simply uses margin loans for living expenses.  The margin loan is only for cash, and is repaid during the recovery.

When has the market been down more than -20% for 13 years?

I don't know how fast the market fell in 1971 - but assume it had dropped 20% by the end of 1972? Then you're taking a loan for 4% each year for the next 12 years, until the value of your investments returns to what it was in 1971 - right?

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #33 on: November 29, 2021, 07:27:29 AM »
https://old.reddit.com/r/FIREUK/comments/r4i2fl/whats_longest_amount_of_time_to_breakeven/

Quote
If I invest into the s&p500 at the peak Price and the price falls and never put anymore money in, what’s the longest amount of time historically has it taken to breakeven?

Quote
Adjusted for inflation, it's about 1971 to 1984.

Don't think you'd want to be carrying a loan for 13 years - with reinvested dividends.
You also don't understand.  At the peak, this approach does nothing.  There's no margin loan.  Only after a 20% crash has already occurred, this approach does not buy or sell assets - it simply uses margin loans for living expenses.  The margin loan is only for cash, and is repaid during the recovery.

When has the market been down more than -20% for 13 years?
I don't know how fast the market fell in 1971 - but assume it had dropped 20% by the end of 1972? Then you're taking a loan for 4% each year for the next 12 years, until the value of your investments returns to what it was in 1971 - right?
Thank you for the reasonable question in this sea of chaos.

The margin loan kicks in after a -20% drop, and once the market is back above that original -20% drop, the margin loan is paid off.  It's 4%/year plus interest.

Portfolio Visualizer can generate results for month to month performance, starting in Jan 1972.  I don't know 1971 performance.  In 1972, the market was -2.45% lower in June & July, so I don't think anything happened then.
https://www.portfoliovisualizer.com/backtest-asset-class-allocation#analysisResults

vand

  • Handlebar Stache
  • *****
  • Posts: 2339
  • Location: UK
Re: Reducing sequence of returns risk with margin
« Reply #34 on: November 29, 2021, 07:33:14 AM »
Isn't there an infamous thread somewhere on Bogleheads that we sometimes link to, which chronicles the ups and downs of someone who tried to invest with leverage throughout the GFC and eventually ended up going bust?

It's car-crash legendary... and probably needs to be (re)read by those who think that using leverage in the manner suggested is a get out of jail free card.

I know that OP is going to say "yeah, but I'm borrowing a tiny amount against the portfolio value." which may be true, but following the strategy doesn't have a circuit breaker other than "hang on until things recover", so borrowing amounts will stack up and compound if we have a sequence of poor years, and before too many years you are suddenly borrowing 20 or 30% of the original portfolio, only the portfolio has now halved or worse so it's more like 40 to 70%

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #35 on: November 29, 2021, 08:19:14 AM »
Isn't there an infamous thread somewhere on Bogleheads that we sometimes link to, which chronicles the ups and downs of someone who tried to invest with leverage throughout the GFC and eventually ended up going bust?
Where in this thread am I talking about "invest with leverage"?

The margin loan is taken out in cash to pay living expenses.  The approach does not buy or sell equities at all.



I know that OP is going to say "yeah, but I'm borrowing a tiny amount against the portfolio value." which may be true, but following the strategy doesn't have a circuit breaker other than "hang on until things recover", so borrowing amounts will stack up and compound if we have a sequence of poor years, and before too many years you are suddenly borrowing 20 or 30% of the original portfolio, only the portfolio has now halved or worse so it's more like 40 to 70%
2 x 30% = 60%, not 70%.

How many 4% withdrawals does it take to add up to 30%?  Is that about 7 years plus interest?

This approach considers a crash a 20% or greater drop, so it does nothing before that.  When the market is back within 20% of it's peak, it pays off the margin loan.  Do you have an example of when the market lost more than 20%, and didn't recover from that point for over 7 years?

With a margin loan to pay for living expenses, those 100% assets after a recovery pay off the 30% margin loan, leaving that person with 70% of their original assets after a 7 year crash.

Consider the flip side: someone who sells their assets for 7 years to pay off living expenses.  Is their portfolio going to last?  They lose 30%, leaving them at 70%... and they keep taking 4% withdrawals for 7 years, dropping them to 42%.  Now their portfolio recovers!  Back to 60% of it's original value.  Their 4% withdrawal rate is now a 6.7% withdrawal rate for the rest of their life.  Is that a better alternative?

Selling stocks in a crash tends to do worse than borrowing cash and paying it back after a recovery.  And this isn't even the full recovery: it's the crash below 20%, and recovery back to that point.  The Great Financial crisis lasted 4.3 years from the smallest drop to complete recovery.  But it was only below 20% for half that time.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #36 on: November 29, 2021, 08:30:39 AM »
I've been thinking about this for a few days and I am not sure how well it would work in practice. I like the idea of it.

A couple of things that I keep going back to. First, I feel like those that would be interested in leveraging in margin would already be leveraged to an extent. I plan on using some, but only up to a point where a large crash would not trigger a margin call.

Assuming that crash came, that's when the time an unleveraged person would leverage margin per your question.

Now what does an already leveraged person do? My assumption here is that you still need the bonds component  likely from another account (Roth IRA for example) to either pay off margin to get away from the call threshold or start pumping into the taxable account to increase liquidity.

So it may prevent you from needing to tent so heavily early on, but still need to tent.

You've probably already said all this but figured I'd chime in.
Before worrying about this topic, you would need a plan to end your margin loan at a loss.  Imagine 2008 or even the great depression - when do you accept the loss and pay off the margin loan?  With no strategy, someone could take -51% losses with 2x margin, and wipe out their account - having zero dollars left.  That's why I don't think your suggestion overlaps mine: I think you should pay down the margin loan and accept losses before you hit -20% with high leverage.

Running my approach through 2008, the deepest part of the crash lasted a few months, followed by about 2 years of recovery.  When assets were lowest, the margin loan was smallest.  When assets had nearly recovered, the margin loan was at it's highest (negative) balance.

If you took leveraged losses and already have a 6% withdrawal rate, I don't know if this approach saves it.

index

  • Pencil Stache
  • ****
  • Posts: 656
Re: Reducing sequence of returns risk with margin
« Reply #37 on: November 29, 2021, 08:49:38 AM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.
How can losses be amplified by an approach that never buys assets?

Your claims of a safe withdrawal rate by definition ignore "sequence of returns risk", which is the whole point of this thread.  A "safe" withdrawal rate can result in going broke if a significant crash occurs very early on.
https://www.kitces.com/blog/understanding-sequence-of-return-risk-safe-withdrawal-rates-bear-market-crashes-and-bad-decades/

Lets say the market falls 30% - trades flat for 2 years then falls an additional 25% in year 3.

Stache: $1.5M
Spending Needs: $60k (4% of $1.5M)
Margin interest - 3.5%

Scenario 1 (borrow on margin):
year 0 - 30% crash - Stache $1.05M - margin loan $60k - interest - $2.1k
year 1 - Stache $1.048M - margin loan $120k - interest -  $4.2k
year 2 - Stache $1.044M - margin loan $180k - interest - $6.3k
year 3 - 25% crash - Stache $776.5k - margin loan $240k - interest - $8.4k (lets say month 6 so half this amount)

IB institutes a loan to value of 1:11 and you are going to get a margin call for around $186k

year 3 - after call -  Stache - ~$586.5k - margin loan - ~$54K   

Scenario 2 (no margin loan):
year 0 - 30% crash - 60k draw - Stache $990k -
year 1 - 60k draw - Stache $930k
year 2 - 60k draw - Stache $870k
year 3 - 25% crash - 60k draw - Stache $592.5k

In the event of a margin call, you are paying back years worth of spending at the new bottom rather than spending those dollars when they were worth 25% more before the second crash.





 

daverobev

  • Magnum Stache
  • ******
  • Posts: 3962
  • Location: France
Re: Reducing sequence of returns risk with margin
« Reply #38 on: November 29, 2021, 09:12:42 AM »

year 3 - 25% crash - Stache $776.5k - margin loan $240k - interest - $8.4k (lets say month 6 so half this amount)

IB institutes a loan to value of 1:11 and you are going to get a margin call for around $186k



Eh, is that right? I thought you could borrow way more than 30% against... oh I guess that's when you borrow to invest?

index

  • Pencil Stache
  • ****
  • Posts: 656
Re: Reducing sequence of returns risk with margin
« Reply #39 on: November 29, 2021, 09:21:15 AM »
Many of your posts seem to have the theme of finding a shortcut to raise investment returns - Crypto, 90% passive 10% active,  factor investing. There is not such thing as a free lunch. Risk can be diversified and you can get lucky by picking the next best sector, factor, or vehicle (crypto), but you equally run the chance of pegging that tilt at its high and underperforming.
The forum rules say to attack an idea, not the person.  This has nothing to do with the current thread - you're simply attacking me.  I've reported you for doing so - please re-read the forum rules.
Did you go on margin at the bottom in March 2020? If not, what makes you think you would pick the bottom during the next large drawdown?
I called the bottom in these forums, and went from 60% to 100% equities on Mar 20 (Friday).  The market bottomed the next trading day, Monday Mar 23.  Once I added margin to my account, I went on margin for the most profitable week in ~35 years, then ended the loan.  The market ignored Covid-19 that I studied, giving me an advantage over the market.

My signature line refers to my Roth IRA investments, which became larger than my total NW from 2019.  My Roth IRA, by itself, more than doubled my NW.  Margin is not allowed in IRA accounts.

Which means I don't need to use margin loans when a crash occurs.  I can ignore a crash, or go on margin.  It won't matter for me.  But for others, sequence of returns risk is a very real problem, and that is also when it's easiest to calculate how margin can be used for living expenses during a crash.

Note also that you're comparing two irrelevant things, because you're too busy focusing on me, instead of the topic.  The topic isn't going on margin to profit in a crash.  The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.

Money is fungible and it really makes no difference if you are spending the money or investing the money. 

The real question of this thread is how deep and how long can a drawdown possibly last and what is the possibility of a double bottom?

If your assumption that a 20%+ crash has a drawdown period of 3 years (first post) is CORRECT, then using leverage to juice returns during accumulation or avoid withdrawals during retirement is ALWAYS the correct answer.

Your assumption relies on calling a market bottom and recovery period - aka market timing. 


frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Reducing sequence of returns risk with margin
« Reply #40 on: November 29, 2021, 09:32:56 AM »
Isn't it extremely likely that the availability of margin loans (or any credit at all) and attractive low rates are going to go away precisely when you most need them to perform this?  And the harsher the crash, the worse those factors will be?


Where in this thread am I talking about "invest with leverage"?

The margin loan is taken out in cash to pay living expenses.  The approach does not buy or sell equities at all.

What's the difference between scenario A and scenario B besides the tax implications?

A. You have $40k invested.  You borrow $40k to fund your spending and keep your original investment invested.
B. You spend your $40k, and borrow a different $40k to invest.

The living expenses are a forgone conclusion. Scenario B you are clearly borrowing money to invest.  You are increasing the amount you otherwise would be able to invest in order to amplify your total return on investment.  You are literally using a loan to invest, this seems like a textbook example of investing with leverage.  Because money is fungible, scenario A is essentially identical (ignoring tax implications). 

You are doing scenario A and claiming you aren't investing with leverage, but the entire scheme seems to be to increase your overall market returns by having more money invested because you've borrowed money. That seems like investing with leverage.

frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Reducing sequence of returns risk with margin
« Reply #41 on: November 29, 2021, 09:37:22 AM »
And also if this could be done successfully during a market crash...why not just do it all the time?  Wouldn't it be even more effective during bull markets when credit is easy and rates are low?  Just get some low rate margin and invest it and profit.

I do this with my mortgage.  I refinanced at the highest possible amount so I could invest the difference.  But my mortgage is safe and not callable. 

index

  • Pencil Stache
  • ****
  • Posts: 656
Re: Reducing sequence of returns risk with margin
« Reply #42 on: November 29, 2021, 09:55:41 AM »
Isn't it extremely likely that the availability of margin loans (or any credit at all) and attractive low rates are going to go away precisely when you most need them to perform this?  And the harsher the crash, the worse those factors will be?


Where in this thread am I talking about "invest with leverage"?

The margin loan is taken out in cash to pay living expenses.  The approach does not buy or sell equities at all.

What's the difference between scenario A and scenario B besides the tax implications?

A. You have $40k invested.  You borrow $40k to fund your spending and keep your original investment invested.
B. You spend your $40k, and borrow a different $40k to invest.

The living expenses are a forgone conclusion. Scenario B you are clearly borrowing money to invest.  You are increasing the amount you otherwise would be able to invest in order to amplify your total return on investment.  You are literally using a loan to invest, this seems like a textbook example of investing with leverage.  Because money is fungible, scenario A is essentially identical (ignoring tax implications). 

You are doing scenario A and claiming you aren't investing with leverage, but the entire scheme seems to be to increase your overall market returns by having more money invested because you've borrowed money. That seems like investing with leverage.

If this was a class assignment, your post would get an "F" grade for plagiarism.

;-)

Money is fungible and it really makes no difference if you are spending the money or investing the money. 

I like your explanation. Money is fungible is a key economic concept that seems like it is easily ignored. There are several threads with thoughts like- "I cashed out the initial principal so it is "free" money", "should I gamble my credit card rewards?", and "This 60% allocation to crypto is all crypto gains so it not really a 60% allocation."

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #43 on: November 29, 2021, 10:25:58 AM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.
How can losses be amplified by an approach that never buys assets?

Your claims of a safe withdrawal rate by definition ignore "sequence of returns risk", which is the whole point of this thread.  A "safe" withdrawal rate can result in going broke if a significant crash occurs very early on.
https://www.kitces.com/blog/understanding-sequence-of-return-risk-safe-withdrawal-rates-bear-market-crashes-and-bad-decades/

Lets say the market falls 30% - trades flat for 2 years then falls an additional 25% in year 3.

Stache: $1.5M
Spending Needs: $60k (4% of $1.5M)
Margin interest - 3.5%

Scenario 1 (borrow on margin):
year 0 - 30% crash - Stache $1.05M - margin loan $60k - interest - $2.1k
year 1 - Stache $1.048M - margin loan $120k - interest -  $4.2k
year 2 - Stache $1.044M - margin loan $180k - interest - $6.3k
year 3 - 25% crash - Stache $776.5k - margin loan $240k - interest - $8.4k (lets say month 6 so half this amount)

IB institutes a loan to value of 1:11 and you are going to get a margin call for around $186k

year 3 - after call -  Stache - ~$586.5k - margin loan - ~$54K   

Scenario 2 (no margin loan):
year 0 - 30% crash - 60k draw - Stache $990k -
year 1 - 60k draw - Stache $930k
year 2 - 60k draw - Stache $870k
year 3 - 25% crash - 60k draw - Stache $592.5k

In the event of a margin call, you are paying back years worth of spending at the new bottom rather than spending those dollars when they were worth 25% more before the second crash.
Two related problems: margin loan interest is not subtracted from assets, it accumulates.  So the balance should be $1.05M until the 25% drop.  And because interest is added to the margin loan balance, it compounds.  The initial $2.1k interest grows to $2.3k over 3 years.

So $1.05M crashes 25% to become $787.5k, and $268k margin loan ($240k + $28k interest).  I see 3x as many assets as debt, and don't see a margin call here.

You are correct to say a margin call collapses this approach: assets are sold at the worst price so far to pay for years of living expenses in a margin loan.  But the margin loan survives this example.

It's also worth comparing this example to the worst crisis in 50 years: the 2008 financial crisis.  That crashed below -20% and recovered in 2.3 years.  The max drawdown was -51%.  The example presented here had a max drawdown of -48%, and spent 4 years under -20%, with no recovery in sight.  So isn't this example harsher than the worst crash in 50 years?

index

  • Pencil Stache
  • ****
  • Posts: 656
Re: Reducing sequence of returns risk with margin
« Reply #44 on: November 29, 2021, 11:28:09 AM »
The topic is taking a margin loan for spending money, and waiting for the stock market to recover.  I gave an example above, showing how it saves money.  The deeper the crash, and the longer it lasts, the bigger the advantage from taking living expenses as a margin loan.
Until you run out of margin and your loan gets called, is the counter argument, right? So you're lowering your losses up to a point, but then they're amplified. You're trading making the core of the distribution narrower for making the tail fatter. Have you run the historical experiment and tested at which withdrawal rates this helps and when it hurts?

Whether it makes sense depends on your risk tolerance, I guess. Since a withdrawal rate that's "safe" already by definition can withstand historical fluctuations, it only makes sense to do this to attempt to sustain an "unsafe" withdrawal rate. So the question is how much extra withdrawal rate it buys you and at which withdrawal rates your margin starts getting called and you're eating pet food under a bridge.
How can losses be amplified by an approach that never buys assets?

Your claims of a safe withdrawal rate by definition ignore "sequence of returns risk", which is the whole point of this thread.  A "safe" withdrawal rate can result in going broke if a significant crash occurs very early on.
https://www.kitces.com/blog/understanding-sequence-of-return-risk-safe-withdrawal-rates-bear-market-crashes-and-bad-decades/

Lets say the market falls 30% - trades flat for 2 years then falls an additional 25% in year 3.

Stache: $1.5M
Spending Needs: $60k (4% of $1.5M)
Margin interest - 3.5%

Scenario 1 (borrow on margin):
year 0 - 30% crash - Stache $1.05M - margin loan $60k - interest - $2.1k
year 1 - Stache $1.048M - margin loan $120k - interest -  $4.2k
year 2 - Stache $1.044M - margin loan $180k - interest - $6.3k
year 3 - 25% crash - Stache $776.5k - margin loan $240k - interest - $8.4k (lets say month 6 so half this amount)

IB institutes a loan to value of 1:11 and you are going to get a margin call for around $186k

year 3 - after call -  Stache - ~$586.5k - margin loan - ~$54K   

Scenario 2 (no margin loan):
year 0 - 30% crash - 60k draw - Stache $990k -
year 1 - 60k draw - Stache $930k
year 2 - 60k draw - Stache $870k
year 3 - 25% crash - 60k draw - Stache $592.5k

In the event of a margin call, you are paying back years worth of spending at the new bottom rather than spending those dollars when they were worth 25% more before the second crash.
Two related problems: margin loan interest is not subtracted from assets, it accumulates.  So the balance should be $1.05M until the 25% drop.  And because interest is added to the margin loan balance, it compounds.  The initial $2.1k interest grows to $2.3k over 3 years.

So $1.05M crashes 25% to become $787.5k, and $268k margin loan ($240k + $28k interest).  I see 3x as many assets as debt, and don't see a margin call here.

You are correct to say a margin call collapses this approach: assets are sold at the worst price so far to pay for years of living expenses in a margin loan.  But the margin loan survives this example.

It's also worth comparing this example to the worst crisis in 50 years: the 2008 financial crisis.  That crashed below -20% and recovered in 2.3 years.  The max drawdown was -51%.  The example presented here had a max drawdown of -48%, and spent 4 years under -20%, with no recovery in sight.  So isn't this example harsher than the worst crash in 50 years?

Margin requirements at IB have been up to 67.5% in recent memory - so with $787.5k you could borrow about 256k. Who the hell knows what the next crash will look like? Japan's crash was much worse than the example above. The FTSE 100 crash from 1999 to 2003 was very similar to the above scenario.

Like I said. Using margin at today's rates, margin requirements, and your assumed worst case scenario is going to win every time. If rates go up, margin requirements go to 75-80%, or there is a double bottom/significant underwater period then this is a losing strategy.

This whole thread is attempting to solve a problem that doesn't really exist if you believe the math behind the Trinity study and the 4% rule. Experiment with some different asset allocations like the golden butterfly and SORR becomes a nonissue. SORR becomes an issue if you are targeting a higher withdrawal rate, looking for a shortcut, as I brought up in my first post on this thread. 

Money is fungible. Your argument is the same as posing the question should you add x% of margin leverage per year the market is 20%+ underwater because stonks go up!

lutorm

  • Pencil Stache
  • ****
  • Posts: 831
  • Location: About the middle of Sweden
Re: Reducing sequence of returns risk with margin
« Reply #45 on: November 29, 2021, 01:19:43 PM »
It would be really awesome if someone could actually run the historical test of this strategy, I'm interested to know whether it would help and by how much.

But one thing that I haven't seen considered: I read the Fidelity terms for margin loans and it clearly states that not only can the interest change by whatever Fidelity feels like at any time, but the margin requirements may as well. So it seems entirely plausible that when the market drops and stuff starts looking sketchy, the brokerage gets cold feet and suddenly increases the margin requirements.

Fundamentally, you're putting your financial fate in the hands of a single other entity (or maybe two or three, if you diversify your loans.) That fact alone is enough for me to not want to touch it with a ten foot pole.

Niceday

  • 5 O'Clock Shadow
  • *
  • Posts: 73
Re: Reducing sequence of returns risk with margin
« Reply #46 on: November 29, 2021, 06:45:27 PM »
During a crisis, margin requirements can change at moment's notice.  The interest rate can also change significantly.  Brokerages will secure their assets and can raise margin requirements significantly and call your loans.  Everybody just wants to survive at that point.  During the internet bubble burst and the mortgage crisis, there was no liquidity.  Heck, during the mortgage crisis, even the money market almost collapsed. Ask people who tried to borrow money then.  There was no money to be borrowed.  There is also psychology at play - you don't really know how long the bear market will last or how bad it'll get.  If you haven't been through the internet bubble and the mortgage crisis, don't let the last 12 years of bull market fool you.  You think 2020 March was bad? It was nothing.

ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 6733
  • Location: A poor and backward Southern state known as minimum wage country
Re: Reducing sequence of returns risk with margin
« Reply #47 on: November 29, 2021, 09:06:22 PM »
ERN was writing about carrying a margin loan from the start of one's retirement. However, the OP by @MustacheAndaHalf proposed a conditional or algorithmic approach - only going into margin AFTER a huge correction or bear market.

I wrote ERN asking if he could retool his spreadsheet to allow for algorithmic approaches. My question was about retiring on a conservative AA and then switching to an aggressive AA when a certain trigger point was reached, such as a xx% drop from the top.
Reading further, I see what you mean.  The article starts out stating something similar to what I mentioned here, but then goes off into borrowing multiples of the portfolio for decades.  Millions are borrowed off a $1 million portfolio, which is unlike my suggestion here.  They did mention the following crash:

From Jan 1973 - Sep 1974 was a -46% correction, with a recovery by Dec 1976.

With current margin rates, margin turns out better in this scenario.  But with 7% treasuries (10 yr) + 6% brokers like to add, 13% margin loan rates would have fared worse than selling assets.

That brings up another point: what is a crash?  I view it as the total percentage loss, and not 10% even if it happens in a few days.  I was thinking somewhere in the 20-25% range, which means I don't consider 1973 to be a crash.  Wikipedia's view:

"There is no numerically specific definition of a stock market crash but the term commonly applies to declines of over 10% in a stock market index over a period of several days."
https://en.wikipedia.org/wiki/Stock_market_crash

Lacking 1973 data, I don't see any mention of sudden percentage drops.  So the 18.2% loss in 1973 was probably not 10% over several days.  If you only go on margin after a 20-25% drop, that starts in 1974 and is 99% recovered in 1975.  So it's more like 2 years on margin, with far lower interest costs.

That's perhaps something I could explore more: what percentage loss is a crash, and so would trigger margin loans (only to pay for expenses).  And what is the breakeven rate on margin loans, where selling assets and the margin loan approach have the same outcome?

Intra-year declines of >20% have occurred every few years for the past couple of decades.
https://intelligent.schwab.com/article/stock-market-corrections-not-uncommon

According to the chart below, a correction >30% can be expected to occur at an average frequency of every 9 years. Of course, we cannot time or predict those dips, and I'm not saying you are trying to. In fact, you are talking about waiting until after the dip has already occurred to make a decision to borrow living expenses rather than sell assets. What we do know is:

1) The stock market will recover some day. It always has in 100% of historical scenarios.
2) The market is safer AFTER a 25% drop than it was BEFORE the 25% drop. Given #1, the bigger the drop, the higher the safety of a decision to buy or hold - even if further losses are ahead. There is simply less room to fall.
3) Bigger drawdowns are rarer than smaller drawdowns.  Across historical scenarios, when the market is down 25%, the odds of it dropping at least another 1% was "X". When it was down 30%, the odds of it dropping at least another 1% was "something less than X", because drawdowns >30% are rarer than drawdowns >25%. So as a drawdown occurs, the likelihood of the market continuing to go down decreases toward zero percent - though we can't see the odds - and then the market must recover.

Thus, by setting an IPS that says borrow living expenses if assets drop XX% in value, you're not market timing or prognosticating, you are bargain shopping for risk using only information available in real time. As prices fall, risk is falling too. When the risk to your remaining assets is low enough, you'd rather borrow living expenses because by holding your remaining assets you are taking a very tiny risk for a very high potential eventual return. I.e. risk assets went on a fire sale that is temporary from the perspective of decades.

I'd suggest setting the "crash" definition a good way down. If you set it too low, such as -15%, you'll be triggered almost immediately into borrowing living expenses. Plus, you will not have dodged much SORR in doing so, because the market could easily fall another 30% from there. Online tools or ERN's spreadsheet can help you identify specific timeframes when your WR would have failed, and you can then make a list of the drawdowns from the previous peak that caused those failures. For example, if the fatal stock drawdowns for your AA were -47%, -35%, and -40% (not actual numbers, haven't looked), you might be inclined to set your trigger at -30% because that's a potentially retirement-killing SORR event and your AA has survived all smaller drawdowns throughout history. Just a suggestion for how to set your crash definition. I know you're meticulous @MustacheAndaHalf , so I fully expect you to look up these numbers and report them here. :)

Looks like no takers on the invitation to rebuild ERNs spreadsheet so we could backtest algorithmic scenarios all day. :(

vand

  • Handlebar Stache
  • *****
  • Posts: 2339
  • Location: UK
Re: Reducing sequence of returns risk with margin
« Reply #48 on: November 30, 2021, 02:49:41 AM »
"If you can't take a small loss, sooner or later you will take the mother of all losses." -- Ed Seykota, trader.

A literal translation in this context - if you cannot bring yourself to voluntarily sell when the market is down, sooner or later you will be forced to sell when the market is much further down.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6659
Re: Reducing sequence of returns risk with margin
« Reply #49 on: November 30, 2021, 06:45:29 AM »
Vand - you have been wrong repeatedly in this thread, and aren't willing to admit it?  You lack evidence for each of your statements, so why continue posting them?  Here's a sampler of your wrong statements in just this thread:


Do you think that in the teeth of the next panic when it seems our way of life is about to change, that someone is going to lend you money at the 1% or 2% rates that are available today? Hell no, borrowing costs go up, up up. If you go broke, they lose their money. So people will want 10% or 15%, or more to risk their capital.
Your claims are false.  In March 2020 the stock market crashed -35%, and after that crash IBKR did not raise margin rates above 2%.  The facts directly contradict your baseless statements.

Isn't there an infamous thread somewhere on Bogleheads that we sometimes link to, which chronicles the ups and downs of someone who tried to invest with leverage throughout the GFC and eventually ended up going bust?
I did not propose "invest with leverage", so you don't understand what I said.  Further, I showed this approach survived the Great Financial Crisis - using evidence, instead of baseless claims.

I know that OP is going to say "yeah, but I'm borrowing a tiny amount against the portfolio value." which may be true, but following the strategy doesn't have a circuit breaker other than "hang on until things recover", so borrowing amounts will stack up and compound if we have a sequence of poor years, and before too many years you are suddenly borrowing 20 or 30% of the original portfolio, only the portfolio has now halved or worse so it's more like 40 to 70%
You got the math wrong here (2 x 30 = 60%, not 70%).  You don't provide any examples or historical data.  "Poor years" usually means substandard returns, not losses.  Significant losses are called a crash, not just poor years.  Years of lower positive returns cause nothing to happen with the approach I suggest.

"If you can't take a small loss, sooner or later you will take the mother of all losses." -- Ed Seykota, trader.

A literal translation in this context - if you cannot bring yourself to voluntarily sell when the market is down, sooner or later you will be forced to sell when the market is much further down.
That is false.  The market was down in 2008, and the approach worked without being "forced to sell", as you claim.

It seems like you don't care if you're wrong over and over.  You don't admit it, you don't correct your mistakes, and you don't have any data to prove your points.