Author Topic: Risk-adjusted exposure to S&P 500  (Read 4275 times)

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Risk-adjusted exposure to S&P 500
« on: October 03, 2018, 02:29:23 PM »
Quote
I believe that people are right that you should mostly be invested in the S&P 500, but totally wrong that you should maintain a consistent allocation like 80/20 or 50/50.

I said this on another thread, not quite realizing how controversial it would be. Not wanting to take the thread further off-topic, I'm trying to spin it off here. The original thread: https://forum.mrmoneymustache.com/investor-alley/advise-me-on-where-to-invest-100k/

I actually feel pretty strongly about this after a lot of thought, and after discussing it with a (very clever) work colleague who recommended I use a particular system to re-allocate my TSP (gov't 401k) monthly. The general reasoning behind doing this is simple: Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move. This is called the Volatility Index (VIX), or "fear & greed index." We can use this info.

Math
When VIX is at 12 (right now), that means that a 1 standard deviation monthly move in the market would be 3.46%. I.e., it would be pretty normal for the market to go up or down 3.46%.

When VIX is at 50 (height of '08-'09), that means that a 1 standard deviation monthly move in the market would be 14.43%.

I am scared of a market where 14.43% moves are normal, because I do not have nerves of steel, and frankly, I do not want to be 100% invested when I can lose 14.43% or more in a month. To use some crude math, I believe that if I should be 100% invested when a 3.46% move is normal, then I should logically be much less invested when a 14%+ move is normal. Ex:  14.43 / 3.46 = 4.17, so 100% / 4.17 = 24% invested. This means that I will continue to only lose 3.46% from my portfolio, even when the market drops another 14.43%.

Good feels
I.e., if I do not want to accept that huge risk by being 100% invested all the time, I do not have to.

This makes me, personally, feel infinitely better about contributing to my account through thick and thin, and I'm much less worried about doing something stupid when the market loses more than half its value again.

Bad for returns?
The worry is that by having a lower stock AA at times, you're hampering your total returns. Now, personally, I'd accept this if it appeared to be true, but the service that I intend to follow has a history of returns going back to 2000, and the cumulative returns for the "Risk Profile" I've chosen (50%) are actually greater than a constant 100% stock allocation. From the performance figures that they posted at this link: https://safer401k.com/how-it-works/performance-since-2000/

Quote
...  the 50% Risk Profile returned 134.98% — quite a bit more than the S&P 500’s 97.5%.

While I do not exactly expect a reply from an $8/mo. service, I am emailing them right now to ask about what the exact inputs are for the risk numbers so I can act more informed about this. My colleague said it is probably the Volatility Index and the 10-year to 2-year bond spread (which is by far the most popular measure of incoming recession). I think that they should make this info available, but I don't see it on the site (it's kind of vague). Regardless, I think that the basic idea behind the approach is very sound, and I intend to follow it totally methodically (I actually stopped lurking and registered to this forum in order to start a journal to keep myself accountable in the beginning).

But really, the basic idea I want to get at here is that it makes sense -- if you know that risk of loss is higher -- to have a bit less money in stocks, thereby protecting yourself from the possibility of large losses. As for myself, the idea has made me a lot more confident that I'm not going to screw up big-time by contributing to my 100% S&P 500 (C Fund) allocation right now, and I'm also not going to be wringing my hands or making stupid, emotional mistakes when the next recession hits. I'll have a plan that I like, and I'll stick to it.

YMMV, no question. But I'm really interested in the theory here, and I'm hoping to feel like less of an idiot than I did in the last thread for thinking this way. The impression I got from my colleague at work is that this is not a controversial idea at all within the professional investment community, and I was expecting (hoping?) to find some sympathizers here.

Thanks!

jacoavluha

  • Stubble
  • **
  • Posts: 232
Re: Risk-adjusted exposure to S&P 500
« Reply #1 on: October 03, 2018, 02:44:36 PM »
Why do you suppose they have chosen to compare the return of their service to the S&P 500 since 2000? I have an idea. Curious what you think.

Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move.

Supporting data?

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #2 on: October 03, 2018, 02:58:32 PM »
Why do you suppose they have chosen to compare the return of their service to the S&P 500 since 2000? I have an idea. Curious what you think.
To show a long enough timeframe to seem relevant? To show what happened in two separate recessionary periods?

Supporting data?
I was taking my friend's word for this one. I assume it to be true because it makes sense if you look at a chart of S&P500 and VIX on Yahoo.

But OK, I Googled. This came up: https://361capital.com/blog-posts/equity-returns-and-the-vix/

Quote
But, what you can say from Graph 7 is that it appears there is a relationship between the current VIX level (the expected future volatility) and the range (or width of the distribution) of future S&P 500 returns, providing more evidence that current expected volatility is predictive of future volatility.



So, there's that.

jacoavluha

  • Stubble
  • **
  • Posts: 232
Re: Risk-adjusted exposure to S&P 500
« Reply #3 on: October 03, 2018, 03:07:34 PM »
The "data" you have shown (can you explain it?), which of course is from a service trying to sell you something, does not prove your statement of, "Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move." Unless you think it does. If so, explain how.

I think they may have chosen the year 2000 because that just happens to coincide with a peak in the S&P 500. And I suspect their data is derived, and not actual. Did "safer401k" exist as an entity in 2000? How about you ask them for returns data from 1997, or 2002.

I do really enjoy the disclaimer on the bottom of their home page though: "Although Safer 401(k) is a very sophisticated tool capable of providing useful calculations it is not designed to replace the advice of a professional investment counselor or your own independent investment research and independent calculations. To rely solely upon the Safer 401(k) tool for investment decisions would be extremely unwise."

marty998

  • Walrus Stache
  • *******
  • Posts: 5828
  • Location: Sydney, Oz
Re: Risk-adjusted exposure to S&P 500
« Reply #4 on: October 03, 2018, 03:15:46 PM »
Looks to me based on that funky VIX graph that there are more dots above 0% than below?

Which makes sense - the market goes up more months than down, and VIX is a lagging indicator - you can only calculate VIX based on events that have already happened.... how often have you see a bounce after a fall? All the time...

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #5 on: October 03, 2018, 03:23:10 PM »
<Moved here from the other thread>

In all sincerity, here's a little gift for you:

https://jlcollinsnh.com/stock-series/

Don't read it now, but do bookmark it, and don't lose it.  I would have given my left arm for this link when I was your age.  Of course, I would have had to wait a couple of decades for the internet to be invented before I could read it, but that's kind of beside the point.

You'll know when it's time for you to read it.  Your future self will thank me. 

erutio

  • Bristles
  • ***
  • Posts: 266
Re: Risk-adjusted exposure to S&P 500
« Reply #6 on: October 03, 2018, 03:31:57 PM »
Unfortunately, I don't think you're going to get a lot of sympathizers here.  What you are talking about is basically market timing, dressed up to sound smart and better than the other "financial advisers" out there.

You don't lose any money as long as you don't sell off when your accounts lose half their value.  The idea is to actually keep buying stocks (ie total stock market fund) at that time to bring your stock AA% back up to your target.  Of course, if you sell off your stocks at the top of the market, then buy all the stocks again a few months later at the bottom of the market, you will be better than the S&P 500 index, but this is attempting to time to market 101.

When you adjust you AA% semi regularly, like one to a handful of times per year, you in essence are already selling a little when the markets are high and buying a little when the markets are low. 

Look, I can come up with a pitch too.  My uncle went to harvard.  He came up with a model to predict how much stock market fund to hold, we call it the Zero to 100% plan.  Base on our model, going back to the year 2000, the model predicted that we would hold 100% total stock market fund until Aug 2000, where it recommends selling all of it, holding cash, then buying 100% total stock market again in March 2003, hold this until Sept 2008, at which time, sell all of it, and buy all of it back in March 2019.  It would have returned >1600% I believe.  Want to buy into our fund?

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 1375
  • Location: Seattle, WA
Re: Risk-adjusted exposure to S&P 500
« Reply #7 on: October 03, 2018, 03:52:23 PM »

YMMV, no question. But I'm really interested in the theory here, and I'm hoping to feel like less of an idiot than I did in the last thread for thinking this way. The impression I got from my colleague at work is that this is not a controversial idea at all within the professional investment community, and I was expecting (hoping?) to find some sympathizers here.

Thanks!

You are doing it wrong.  The first thing you should do before considering any investment strategy is see what the haters are saying.   Cheerleaders are not your friends.  You want people to poke holes in it, tell you what you don't know.   I don't know how their black box model works (should be a big red flag right there), but I do know it is dead-ass simple to backtest a strategy that beats the S&P over the last 10 or 20 years.  Create one that beats it over the 30 years is really hard.  I'm willing to bet a bottle of beer that this strategy doesn't work if you backtest it farther.   

One thing that seems scary to me is they somehow seem to think that volatility is equal to risk.  That's not true, and I strongly suspect that's why they decided to start their return sequence at 2000.  Because in the 1990s the market was very volatile--in an upward direction.   Capping volatility means capping returns.  Since the stock market generally goes up, you want to be careful tamping down the volatility because you are mostly tamping down your gains. 

Another thing that should give you pause is the R2 in chart you posted shows there is almost no (maybe no) correlation between volatility and stock market returns.  So why is it in the model? 

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #8 on: October 03, 2018, 03:58:46 PM »
The "data" you have shown (can you explain it?), which of course is from a service trying to sell you something, does not prove your statement of, "Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move." Unless you think it does. If so, explain how.
Look at the chart. When VIX is higher, the market moves more. The market has never lost more than 3% when VIX is 10. The market has lost 17% when VIX is 40. Big difference.

VIX is a lagging indicator
No, VIX is not a lagging indicator. It's derived from current option prices. So it's the combined opinion of option traders on how much the S&P 500 might move.

Unfortunately, I don't think you're going to get a lot of sympathizers here.  What you are talking about is basically market timing, dressed up to sound smart and better than the other "financial advisers" out there.

You don't lose any money as long as you don't sell off when your accounts lose half their value.  The idea is to actually keep buying stocks (ie total stock market fund) at that time to bring your stock AA% back up to your target.  Of course, if you sell off your stocks at the top of the market, then buy all the stocks again a few months later at the bottom of the market, you will be better than the S&P 500 index, but this is attempting to time to market 101.
I get that you don't like things that sound like market-timing, but this approach is literally doing the opposite of your example. This portfolio would sell when the market goes down. It has nothing to do with "calling the top" or "catching a falling knife" or any of that garbage. What I'm talking about is seeing a reliable indicator that says "you can lose 14%" and choosing to NOT lose 14% by re-allocating.

You are doing it wrong.  The first thing you should do before considering any investment strategy is see what the haters are saying.
I guess that's why I'm here, huh? This place is apparently pretty dogmatic.

One thing that seems scary to me is they somehow seem to think that volatility is equal to risk.
If you're 100% invested in the S&P 500 and you stand to lose 14% this month based on volatility estimates, then I think it's fair to at least call volatility a type of risk, no?

MDM

  • Walrus Stache
  • *******
  • Posts: 8775
Re: Risk-adjusted exposure to S&P 500
« Reply #9 on: October 03, 2018, 04:25:43 PM »
If you're 100% invested in the S&P 500 and you stand to lose 14% this month based on volatility estimates, then I think it's fair to at least call volatility a type of risk, no?
Why are you focused on the "stand to lose" component, when per the chart you showed when the VIX is ~40-50, next month gains outnumbered losses 6 to 3?

jacoavluha

  • Stubble
  • **
  • Posts: 232
Re: Risk-adjusted exposure to S&P 500
« Reply #10 on: October 03, 2018, 05:07:59 PM »
The "data" you have shown (can you explain it?), which of course is from a service trying to sell you something, does not prove your statement of, "Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move." Unless you think it does. If so, explain how.
Look at the chart. When VIX is higher, the market moves more. The market has never lost more than 3% when VIX is 10. The market has lost 17% when VIX is 40. Big difference.

I still don’t see the “predict with great accuracy” part.

And where are all the actively managed funds using this strategy and outperforming?


wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #11 on: October 03, 2018, 05:28:16 PM »
Why are you focused on the "stand to lose" component, when per the chart you showed when the VIX is ~40-50, next month gains outnumbered losses 6 to 3?
That's a really good question, and I think that this demonstrates exactly why a buy-and-hold strategy works so well -- volatility doesn't change the fact that the market, in the long run, is going to work in your favor. There's no question that by reducing exposure when there's a lot of volatility, you're missing out on those +10% months and turning them into +4% months instead.

I still don’t see the “predict with great accuracy” part.

And where are all the actively managed funds using this strategy and outperforming?
Standard deviation of monthly returns goes up in nearly-perfect linear fashion with VIX. Current volatility predicts future volatility, regardless of whether that means "up" or "down." I don't know how else to explain that. I guess I can try running some numbers myself to demonstrate, but...

And I mentioned this in the last thread, but again, no large fund is interested in absolute returns alone. Risk-management is first and foremost. The reason a fund would use a volatility-targeting approach is to reduce the volatility of the portfolio, which is considered a good thing in its own right, since it means you end up compounding returns more reliably -- no huge drawdowns in portfolio value. (Rule #1, don't lose money.)

PizzaSteve

  • Pencil Stache
  • ****
  • Posts: 505
Re: Risk-adjusted exposure to S&P 500
« Reply #12 on: October 03, 2018, 05:33:26 PM »
The "data" you have shown (can you explain it?), which of course is from a service trying to sell you something, does not prove your statement of, "Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move." Unless you think it does. If so, explain how.
Look at the chart. When VIX is higher, the market moves more. The market has never lost more than 3% when VIX is 10. The market has lost 17% when VIX is 40. Big difference.

I still don’t see the “predict with great accuracy” part.

And where are all the actively managed funds using this strategy and outperforming?
The VIX is a backwards looking measure.  It has no predictive value.  It is also skewed by large market drops, so most of the differences in trends are attributable to a few big drops. 

So you are saying that in the months with big corrections, volatility was high. So basically you uncovered the acxion that markets fall quickly, grow more steadily.  This is absolutely useless as a forward looking tool.  Yes, after black Friday is over VIX is suddenly high.  What do you do? You dont know if the market will fall further (it did), nor when it will recover.  All you can do is get out of the market and limit gains or losses, but time out of the market is lower returns overall, on average.  It is an old approach, but not a winning one.  Buy and hold beats it.

<removing content and withdrawing from the dialog...dont appretiate tone of OP...peace and good luck>
« Last Edit: October 03, 2018, 05:59:25 PM by PizzaSteve »

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #13 on: October 03, 2018, 05:40:47 PM »
I am interested in the qualifications of your `smart friend.'  Have some pretty strong credentials myself and am curious how this idea was developed and vetted.  I am a pretty big fan if the efficient market hypothesis and dont read anything convincing so far.
My colleague is a programmer. He trades index options with an automated strategy. Some of his friends ended up in the finance industry. We get lunch and talk a lot. If you have questions for him, I'll be sure to ask. He knows a lot more than the rest of us.

The VIX is a backwards looking measure.  It has no predictive value.
This is demonstrably false. Which makes me question your "strong credentials." VIX is derived from options that expire in one month. One month in the future.

You could conceivably argue that VIX is only predictive of future volatility to the extent that current volatility is predictive of future volatility -- but that's not even up for debate. Quoting the linked study again...

Quote
But, what you can say from Graph 7 is that it appears there is a relationship between the current VIX level (the expected future volatility) and the range (or width of the distribution) of future S&P 500 returns, providing more evidence that current expected volatility is predictive of future volatility.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #14 on: October 03, 2018, 05:46:50 PM »

And I mentioned this in the last thread, but again, no large fund is interested in absolute returns alone. Risk-management is first and foremost. The reason a fund would use a volatility-targeting approach is to reduce the volatility of the portfolio, . . .

You're confusing the method for the reason. The reason fund managers need to manage volatility is that their investors could pull their money out of the fund at any time.  A run on the fund could force them to liquidate a large chunk of their fund at market lows if they didn't do that.  Managing volatility is the method they employ to do that, not the reason that they do that.

Are you managing other people's money?  Or are you managing your own money?  If the latter, then the only one who can initiate a run on your fund is you.

. . . no huge drawdowns in portfolio value. (Rule #1, don't lose money.)

You keep repeating that apparently without understanding what it means, fundamentally.

Temporary draw downs in value does not equal losing money.  Selling off part of a position that has a capital gain is a taxable event.  The capital gains taxes you pay will never be compounding for you again.  That's a permanent future pay cut.  That's losing money.   The more often you do it, the more money you will lose.

You cite Buffet's rule #1, as if it were somehow an argument in favor of your programmer buddy's approach, apparently without understanding what it means.

Fund managers have a much bigger set of problems to deal with than you, as an individual investor has.  That gives you an advantage over them.  But if the only tool you pull out of the fund manager's toolbox is a hammer, then every problem you face will look like a nail; even when it's not a nail.

I'll repeat this, because it bears repeating: a temporary loss of value does not equal losing money.
« Last Edit: October 03, 2018, 08:01:58 PM by ILikeDividends »

PizzaSteve

  • Pencil Stache
  • ****
  • Posts: 505
Re: Risk-adjusted exposure to S&P 500
« Reply #15 on: October 03, 2018, 05:46:58 PM »
For the curious, VIX is a volatility index measure.  People create derivative financial products and trade them based on it.  Its like the S&P 500 index has futures derivative products that can be traded, but the S&P500 index itself measures the weighted price of a basket of specific stocks.

https://www.investopedia.com/terms/v/vix.asp  It attempts to be forward looking using data.

About 2 weeks ago I attended a small meeting with this guy at Charles Schwab https://www.schwab.com/resource-center/author/jeffrey-kleintop.  He is the one who explained how average VIX is less useful than median VIX, because a few days skew the average high, and also how it is not predictive.  So if you dont believe me (I do go out there and stay educated still, ask him, as he was my source for the comment).

<edited down to useful content>
« Last Edit: October 04, 2018, 09:17:21 AM by PizzaSteve »

markbike528CBX

  • Pencil Stache
  • ****
  • Posts: 836
  • Location: the Everbrown part of the Evergreen State (WA)
Re: Risk-adjusted exposure to S&P 500
« Reply #16 on: October 03, 2018, 05:48:26 PM »
Per the "Top is IN" thread. First post.

Fear is back, VIX above 15, XIV breaking down. SPY to follow, earnings will be a reality check.

Note that the S&P and most US indices are WAY up 25% since April 2017

If thorstach our Beloved Leader is this far off, how can mere mortals such as us have hope of doing better.

Sounds like someone is selling snake oil.

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #17 on: October 03, 2018, 06:02:21 PM »
You cite Buffet's rule #1, as if it were somehow an argument in favor of your planned approach, apparently without understanding what it means.

I'll repeat this, because it bears repeating: losing value does not equal losing money.
Fine, but "losing value" is still the opposite of compounding, and I want to compound.

So my 30 plus years working at high levels in finance (advised Fortune 50 CEOs, CFOs for real money) and business degrees from the top US finance b school doesn't count vs a programmer for a trading firm?

Please read this. [...]

https://www.investopedia.com/terms/v/vix.asp

Do what you like though.  I will delete my replies, since you are sensitive.
From the article you linked:

Quote
The Volatility Index, or VIX, is an index created by the Chicago Board Options Exchange (CBOE), which shows the market's expectation of 30-day volatility. It is constructed using the implied volatilities on S&P 500 index options. This volatility is meant to be forward looking, is calculated from both calls and puts, and is a widely used measure of market risk.

I said my colleague knows more than us because I think he does -- not to offend you or your degrees. VIX is forward-looking, though.

jacoavluha

  • Stubble
  • **
  • Posts: 232
Re: Risk-adjusted exposure to S&P 500
« Reply #18 on: October 03, 2018, 06:07:00 PM »
If it’s forward looking then let’s see the outperformance compared vs SP500, not starting from a market peak like 2000

PizzaSteve

  • Pencil Stache
  • ****
  • Posts: 505
Re: Risk-adjusted exposure to S&P 500
« Reply #19 on: October 03, 2018, 06:08:34 PM »
<edited down to useful content>

The words 'meant to be (based on futures and options pricing)' are the key words.  The superbowl winner is meant to be predictive of a bull or bear stock market as well.  Are you going to bet your future on that?

I am the first to admit I dont know everything and might missuse a term, but please...your challenging tone is not helpful.
« Last Edit: October 04, 2018, 09:19:11 AM by PizzaSteve »

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #20 on: October 03, 2018, 06:13:45 PM »
You cite Buffet's rule #1, as if it were somehow an argument in favor of your planned approach, apparently without understanding what it means.

I'll repeat this, because it bears repeating: losing value does not equal losing money.
Fine, but "losing value" is still the opposite of compounding, and I want to compound.
Wow.  How do you come up with these through-the-looking-glass conclusions?  Of course you're still compounding during a market melt down.  It's temporary.  You'll still have more of those shares when the market recovers.

While the market is low, your dividends will still be paid and reinvested at a more attractive price.  And for the next 30 years, those shares you bought with those dividends will pay dividends, and so on.  That's how compounding works.  And at your age, while you're still working, you'll still be shoveling money from every paycheck into a more attractively priced market.  At least you should be.

As a side note: I'm glad you got that link I gave you.  However, I suggested that you not read it right now for a reason.  I don't think you're open enough to new ideas that conflict with your programmer pal's way of thinking; not yet.  Heck, I wasn't very open to opposing viewpoints when I was your age.  The cockiness of youth, and all that.  I get it.

Like me, you're probably going to have to make every market blunder known to man a couple of times, at least, before you mellow enough to accept new ideas.  So go do that.  I had to.  You're on the right track to learn what not to do all on your own.  So is your programmer buddy.  I can tell you from personal experience, that is very valuable experience to have; expensive, but truly valuable. Set that link aside and revisit it in about 10 years.  You'll still have time to grow a nice little nest-egg before you retire.

For what it's worth, I was a very well paid programmer for 20 years.  Machine language (PDP-8/Assembly), COBOL, Fortran, Pascal, Prolog, Basic, C, C++, C sharp, Java, HTML, PHP, CS, PL/SQL; you name it, I've been paid as a professional to write all of those languages on a plethora of different operating systems, most of which are now obsolete. 

I'm sure I've left off a few languages from the list, but you get my point, eh?  I can learn a new language tomorrow and code circles around just about anyone else in about a week, and still have time to help them debug their own code in my spare time.  My talent for programming, however--and unfortunately--didn't cross over into an inherent talent for finances.  I had to learn that the hard way.  Word to the wise.
« Last Edit: October 03, 2018, 11:39:29 PM by ILikeDividends »

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #21 on: October 03, 2018, 06:50:25 PM »
OK.

I have, on this thread and in PMs, been advised to exercise caution.

Since my plan is currently to use this system as a long-term solution to my retirement savings, I probably owe it to myself to understand it better than I do. Fortunately, I am certain that my colleague understands it, and I am hopeful that I will receive more information from the people at Safer 401(k).

I will update as I learn more.

You can't get rid of me that quickly.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #22 on: October 03, 2018, 06:53:23 PM »
You can't get rid of me that quickly.

No one's trying to get rid of you.  Hang around for awhile.  You might just get to retire early after all.  Then it will be your turn to help someone younger retire early too.

;)
« Last Edit: October 03, 2018, 07:55:53 PM by ILikeDividends »

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #23 on: October 03, 2018, 07:59:13 PM »
No one's trying to get rid of you.  Hang around for awhile.  You might just get to retire early after all.

;)
I'm optimistic!

Radagast

  • Handlebar Stache
  • *****
  • Posts: 1263
  • Location: West of the Mountains, East of the Sea
Re: Risk-adjusted exposure to S&P 500
« Reply #24 on: October 03, 2018, 08:02:22 PM »
I said my colleague knows more than us because I think he does -- not to offend you or your degrees. VIX is forward-looking, though.
WOW YOUR COLLEAGUE PROGRAMMED A TIME MACHINE!!!!!!!1 hE must be very smart and rich. You should definitely do what he says.

Radagast

  • Handlebar Stache
  • *****
  • Posts: 1263
  • Location: West of the Mountains, East of the Sea
Re: Risk-adjusted exposure to S&P 500
« Reply #25 on: October 03, 2018, 08:02:38 PM »
But seriously, if it was a great idea, every hedge fund and trading desk would do it. Then there would be trillions of dollars pursuing the strategy. Then it would fail epically because trillions of dollars were following it. So most likely, it either 1) does not work, or 2) is about to fail epically.

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #26 on: October 03, 2018, 08:10:22 PM »
WOW YOUR COLLEAGUE PROGRAMMED A TIME MACHINE!!!!!!!1

But seriously, if it was a great idea, every hedge fund and trading desk would do it. [...] So most likely, it either 1) does not work, or 2) is about to fail epically.

Thanks. Very helpful.

If anyone else is interested, the basic concept that I'm talking about here is a lot like "risk-parity," which has been around for a long time as part of Modern Portfolio Theory. Some people think it's good, some people don't.

From a Forbes article (https://www.forbes.com/forbes/2012/0507/investing-risk-stocks-wealth-wringing-volatility-out-of-returns.html):
Quote
Partridge's solution is a strategy called risk parity, a fancy name for selecting investments based not on their returns but on their volatility. Imagine two simple portfolios, one with 2% volatility and the other with 4% volatility. Both have 8% expected returns, but the more volatile portfolio is twice as likely to fall hard and dig itself into a hole it can't get back out of.

Hopefully this helps you guys realize that I'm either not taking crazy pills -- or at least I'm not the only one taking crazy pills out there.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #27 on: October 03, 2018, 08:21:37 PM »
Forbes is only one of many drug dealers peddling crazy pills.  That's how they make their money.  They can be entertaining, and fairly safe to ingest, but only if you understand that one key point.

If you're prone to actually acting on their publications or broadcasts, then it would be safer to just avoid them altogether.  They are the primary source of any "feels" panic you might be likely to act on in a panic.  They get paid to add fuel to the fire.

A steady diet of them is kind of like overdosing on recreational paranoia; legal in all fifty states, but not necessarily good for you.
« Last Edit: October 03, 2018, 08:32:42 PM by ILikeDividends »

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #28 on: October 03, 2018, 08:33:47 PM »
Forbes is only one of many drug dealers peddling crazy pills.  That's how they make their money.  They can be entertaining, and fairly safe to ingest, but only if you understand that one key point.

If you're prone to actually acting on their publications or broadcasts, then it would be safer to just avoid them altogether.
Geez.

OK. Bridgewater, quite possibly the world's most influential hedge fund, with $160bn AUM, closed to new investment, 25% yoy returns in 2001 - 2010, etc... does a lot of risk-parity.

Ray Dalio is not crazy. Neither is risk-parity.

https://www.bridgewater.com/research-library/risk-parity/

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #29 on: October 03, 2018, 08:41:48 PM »
Good for Bridgewater.  What are their YOY returns over the last 100 years?

Dale Earnhardt Jr. also became rich racing (I just knew I'd find a way to make him relevant to this thread).

Both approaches are perfectly reasonable for the few people who have the aptitude and skill set to do it.

I can name names of basketball players who became filthy rich playing that game.  That doesn't mean I can.

This forum is populated by people that don't want to (or more likely are unable) to excel at any of those professions, but who still want to retire early; some who have already, and others who are already on a glide path to retire early, with the least effort possible. 

Yeah, I'm lazy.  Guilty as charged.  But I stopped working in 2010 just so that I could fully express my laziness to my heart's content.  I start collecting social security in about 6 months.  I won't even have to draw down my stash after that kicks in; which, by the way, is bigger now than it was in 2010.   I get my medical from the VA, so I couldn't care less what happens to Obama-care.  You haven't even begun to see lazy from me yet.  My best, and laziest days are still ahead of me.
« Last Edit: October 03, 2018, 09:26:20 PM by ILikeDividends »

Radagast

  • Handlebar Stache
  • *****
  • Posts: 1263
  • Location: West of the Mountains, East of the Sea
Re: Risk-adjusted exposure to S&P 500
« Reply #30 on: October 03, 2018, 09:00:54 PM »
Recommended reading: William Bernstein "The Four Pillars of Investing" and "Skating Where the Puck Was"

Lets take Forbes as our gospel. There is really no such thing as an enduring low risk high return investment, especially not one that can be exploited indefinitely by somebody recreating at a computer. If an investment that has high returns, low volatility, and is poorly correlated with other financial assets is discovered, the very first person to discover it will make out like a bandit. Every other person will rush in to take advantage of it, further increasing the returns of the original investor and lowering the returns of all subsequent investors. High return cannot exist for long in combination with either low correlation or low volatility, or especially both. Promises of that are actually an investing warning sign. Dalio (and Swenson, and probable Harry Browne) was a pioneer who reaped massive profits when the herd followed and drove his assets up in price. People starting after 2010? A lot higher hill to climb. Further, hedge funds in aggregate have been showing continuously reduced performance, to the point they under perform a combination of stocks and treasury bonds.

...long discussion. Anyhow, do what's best for you. As long as you don't use leverage or certain options you shouldn't lose too much money...

Systems101

  • Stubble
  • **
  • Posts: 149
Re: Risk-adjusted exposure to S&P 500
« Reply #31 on: October 03, 2018, 09:50:03 PM »
I'm going to add to the chilly reception.  The data you refer to is narrow in time frame, and longer term testing doesn't support the investing style you are advocating.  You are also using terms in ways that indicate you haven't done your research.

Well-informed traders have historically been able to predict, with great accuracy, how much the S&P 500 will move.

Traders make financial bets as to how the market will move.  In the aggregate, buyers of options are losing money, so those that are fearful or trying to capture big moves are more wrong than right (see: https://www.investopedia.com/articles/optioninvestor/03/100103.asp )

"Great Accuracy" is a wild claim.  The R-squared of the VIX (implied volatility) predicting the absolute value of actual volatility (as a percentage) for the next month is in the 20% range.  It's higher than I expected, but it's certainly not "great".  Also (thinking out loud), how much of that is covariant with the momentum factor?

I believe that if I should be 100% invested when a 3.46% move is normal, then I should logically be much less invested when a 14%+ move is normal.

Lots of people once held a belief that the world is flat. Logically, if you sailed to the end of the world, you would fall off.  However, beliefs are not always correct.  The response you are receiving may be from folks who actually do analyze data - some of us for a living - and we test the theories with real data to make sure our beliefs are correct.  Often, we find they aren't.

So, let's test your logic.  Let's follow a simple algorithm: Investment rate is min(100,115-VIX).  So if VIX is under 15, we are fully invested, then if it's above 15, we scale down in a linear fashion.  This should give a flavor of whether your belief is correct.  We invest for a month at a time.

If we select from the top of the market in 2000, then our return is about 2% above the S&P 500.  If I reset the start date to January 1993, then we are about 9% behind the S&P 500.

The average monthly return fully invested is 0.856%, with the modified logic it's 0.811%.  The standard deviation of return is 4.1% vs 3.7%.  So the volatility went down, but so did the return.  Since folks here are maximizing return and accepting the volatility, I don't know why anyone here would be interested in using the VIX as you describe.  You will note I don't use the word "risk" - because volatility is not a good proxy for risk in the real world (it's just used that way as a simplification in academic papers... to quote Radagast, this is a "long discussion"...)

You can pull up ^VIX and SPY and the monthly historical data on Yahoo to build your own data or reproduce what I have summarized.

The biggest criticism of my analysis here is that it *still* fails to include some significant historical periods (pre-1990s) which have very different market dynamics.  Unfortunately, the concept of VIX and options don't go back far enough, at least in data I used.

There's no question that by reducing exposure when there's a lot of volatility, you're missing out on those +10% months and turning them into +4% months instead.

The problem is that then you lose against staying fully invested as the summary data I showed above indicates.  You're failing to participate in needed upside.

VIX is forward-looking, though.

This is one of those statements that makes my hair stand up on end.  It's pedantically correct, but the statement doesn't actually address the issue or move the discussion forward.  Specifically, VIX is forward looking in that it represents *implied* volatility.  People have made real bets with real money... But that is a useless statement.  What matters is whether it's forward-looking *indicator of performance*.  It is not.

My saying a hurricane will strike Idaho tomorrow is forward-looking, but useless (since it's a fabrication).  What you see on CNBC is forward-looking, and I think it's useless, too.  As you can see in the article you linked, the VIX is *not* predictive of return.  It is actually a *coincident* indicator of stock performance based on that article.  So it is not useful for predicting return, and thus shouldn't be used for asset allocation.

Hopefully this helps you guys realize that I'm either not taking crazy pills -- or at least I'm not the only one taking crazy pills out there.

Except this only reinforces the problems I expressed above.  See Key Observations #9 and #11: http://orcamgroup.com/wp-content/uploads/2013/08/TheRiskParityApproachtoAssetAllocation2010.pdf

The reality is that risk parity outperformed since 2000, but massively underperformed in the 1990s.  Your Bridgewater example will suffer the same problem.

The folks here are playing a statistical long game based on decades of data and lots of analysis.  There is a massive dataset that shows passive investing is - statistically - the way to go.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #32 on: October 03, 2018, 10:57:19 PM »
The folks here are playing a statistical long game based on decades of data and lots of analysis.  There is a massive dataset that shows passive investing is - statistically - the way to go.
I actually understood this part of your post.  ;)

+100

Great post!

(I understood more than that; I just wanted to express my appreciation)
« Last Edit: October 03, 2018, 11:15:27 PM by ILikeDividends »

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #33 on: October 04, 2018, 12:04:48 AM »
So, let's test your logic.  Let's follow a simple algorithm: Investment rate is min(100,115-VIX).  So if VIX is under 15, we are fully invested, then if it's above 15, we scale down in a linear fashion.  This should give a flavor of whether your belief is correct.  We invest for a month at a time.

The average monthly return fully invested is 0.856%, with the modified logic it's 0.811%.  The standard deviation of return is 4.1% vs 3.7%.  So the volatility went down, but so did the return.
But this is exactly what I'm hoping to achieve. Less start-date sensitively, less liquidation/draw timing sensitivity, less worry. That's why I picked the 50% loss limit. The system optimizes for that level of risk tolerance. If I picked a 95% loss limit, I'd be 100% in all the time just like you said, even in the worst of times -- but I didn't do that.

So you're basically telling me that I'm stupid (I can't use terminology correctly, I believe that the "world is flat," I'm wrong about VIX), but then you're proving to me that a cursory analysis of adjusting positions using VIX actually lowers portfolio volatility. Meaning that VIX clearly has some predictive value, and the Safer 401(k) method of scaling risk up and down according to volatility is in fact likely to work.

Why, after that, are you still telling me that VIX is useless garbage?

Since folks here are maximizing return and accepting the volatility, I don't know why anyone here would be interested in using the VIX as you describe.
If everyone here literally has 100% of their liquid assets in stock indexes all the time, then sure, you're right. Is anyone here actually doing that in earnest? Because I'm not, and I need a smart portfolio-allocation solution for the fact that I'm not doing it.

What started this thread:
Quote
I believe that people are right that you should mostly be invested in the S&P 500, but totally wrong that you should maintain a consistent allocation like 80/20 or 50/50.

You're making me more confident that this is all a good idea.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #34 on: October 04, 2018, 12:14:47 AM »
You can reduce volatility very simply by allocating a greater percentage to cash equivalents or fixed income.  No one is suggesting that you should take on more equity risk than you are comfortable in taking on.  You don't need to twist yourself into knots over unproven, and misunderstood schemes, in order to do that.

I'm not calling you stupid; at least no more stupid than I was at your age. And I don't think others are calling you stupid either.  Spectacularly misinformed or uninformed, hard-headed and stubborn, yeah, there's a bit of that kind of mockery going on in this thread.  You should admit, though, if you're going to be honest with yourself, the mockery is well deserved.  Man up.  Ultimately, it means nothing.

Genius flourishes by learning from mistakes--your mistakes and other's mistakes--not by what you think you might know in the moment.
« Last Edit: October 04, 2018, 02:48:14 AM by ILikeDividends »

jmswtc

  • 5 O'Clock Shadow
  • *
  • Posts: 9
Re: Risk-adjusted exposure to S&P 500
« Reply #35 on: October 04, 2018, 09:07:12 AM »
I guess here is the problem I am seeing with this (along with all the others people have pointed out).


You said yourself you will only log on once a month to reallocate (and no more than that). Lets say the first of the month for the purpose of this exercise. So, you log on October 1st and the VIX is 13 (which means 100% invested in stocks based off what I've read). However, something crazy happens and the Vix jumps to 30 by October 5th and markets are starting to go down.


You said yourself that you are only logging on once a month so you are stuck in your 100% stock allocation for the next 25 days until you reallocate on November 1st.


It just seems odd that you are admitting to taking this "market timing" approach, and yet you are only adjusting your allocation based on what the VIX index is on only 12 days a year. That just seems contradicting to me since the Vix can jump all over the place. What if it's 30 on the first day of the month, but 5 days later it's back at 13 and stays there the rest of the month?

Maybe I'm thinking about this wrong because I generally don't agree with it in principle, but I have no problem with you doing it. It's your hard earned money and I'm not going to be the one to tell you not to do it. I'm just pointing out the ways I'd be uncomfortable with it.
« Last Edit: October 04, 2018, 09:08:55 AM by jmswtc »

PizzaSteve

  • Pencil Stache
  • ****
  • Posts: 505
Re: Risk-adjusted exposure to S&P 500
« Reply #36 on: October 04, 2018, 09:32:54 AM »
You can reduce volatility very simply by allocating a greater percentage to cash equivalents or fixed income.  No one is suggesting that you should take on more equity risk than you are comfortable in taking on.  You don't need to twist yourself into knots over unproven, and misunderstood schemes, in order to do that.

I'm not calling you stupid; at least no more stupid than I was at your age. And I don't think others are calling you stupid either.  Spectacularly misinformed or uninformed, hard-headed and stubborn, yeah, there's a bit of that kind of mockery going on in this thread.  You should admit, though, if you're going to be honest with yourself, the mockery is well deserved.  Man up.  Ultimately, it means nothing.

Genius flourishes by learning from mistakes--your mistakes and other's mistakes--not by what you think you might know in the moment.
Well said.  When someone doubts and mocks people offering to help (for free), who are stated they are efficient market investors with a market timing plan, they will take some hits.

But again, its his life,  As you said, we are retired with our millions already.  On the plus side, this should only reduce returns by the percentage of time out of the market, less transaction and advisory costs.  If he saves well he should be fine.
« Last Edit: October 04, 2018, 09:59:05 AM by PizzaSteve »

jacoavluha

  • Stubble
  • **
  • Posts: 232
Re: Risk-adjusted exposure to S&P 500
« Reply #37 on: October 04, 2018, 09:57:52 AM »
So the volatility went down, but so did the return.
But this is exactly what I'm hoping to achieve.
How about an apples to apples comparison then. Find the static SP500:Fixed income portfolio that matches that volatility. Then compare the returns to the variable allocation portfolio that you're advocating. I'm genuinely curious what the results show.

PizzaSteve

  • Pencil Stache
  • ****
  • Posts: 505
Re: Risk-adjusted exposure to S&P 500
« Reply #38 on: October 04, 2018, 10:04:34 AM »
So the volatility went down, but so did the return.
But this is exactly what I'm hoping to achieve.
How about an apples to apples comparison then. Find the static SP500:Fixed income portfolio that matches that volatility. Then compare the returns to the variable allocation portfolio that you're advocating. I'm genuinely curious what the results show.
In theory, that is hard to do because the VIX based plan will have variable returns, depending on market volatility, so it would be a range of returns similar to a high of 100% stock (markets unusually calm like this year so far) to a low of 50% stock (using OP example), of course less transaction and advisory costs.  Your stock portfolio would earn avg time in market x market returns, your other chunk would earn average time in asset class x average returns for that asset (say bonds or cash), but its like having an unknown allocation and picking one at random (in theory).

If you believe VIX can predict the future, skys the limit.  Market timing works and you outperform.

One can back test, but that is only backwards looking.  As my Schwab guy said, VIX is highly skewed by a few single day and even intraday shock events.  Its numbers must be carefully considered since something like a terrorism scare and one day blip will bump it up for a while.

PS. Sorry to jump back in, cant help myself sometimes.  If you want to make money in trading and market timing I would advise researching less efficient markets. 
« Last Edit: October 04, 2018, 02:53:57 PM by PizzaSteve »

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #39 on: October 04, 2018, 10:21:00 AM »
You cannot logically think this ...

You can reduce volatility very simply by allocating a greater percentage to cash equivalents or fixed income.

... and this ...

Since folks here are maximizing return and accepting the volatility, I don't know why anyone here would be interested in using the VIX as you describe.
... at the same time. They are at odds.

Either @Systems101 is right that we should all be 100% invested in stock indexes and accepting all the blows all the time, or what I'm proposing is a completely legitimate way to lower portfolio volatility while maximizing stock exposure.

What I'm saying is that if you do want to lower your portfolio volatility, adjusting according to stock volatility is a much better way since it allows you to be 100% invested the majority of the time instead of 80/20 or 60/40 or whatever you might do to lower volatility otherwise. The 10-year performance of my 50% Risk Profile portfolio is nearly identical to a 100% buy-and-hold over the period, and so it will clearly beat a "more conservative allocation," which would have been dragged down by fixed-income for a whole decade.

I guess here is the problem I am seeing with this (along with all the others people have pointed out).
...
You said yourself that you are only logging on once a month so you are stuck in your 100% stock allocation for the next 25 days until you reallocate on November 1st.
I don't really see this as a problem, though. I know a lot of folks are trying to characterize it this way, but I'm not trying to "time the market" -- I'm trying to use a practical system that's better than 80/20. Right now, I'm 100% invested in stocks, and that's not likely to change soon. When it does, I'll readjust, but I don't feel the need to monitor my account every day -- and I really don't want to.

If you believe VIX can predict the future, skys the limit.  Market timing works and you outperform.
I think this is a fairly reasonable characterization of the discussion so far, but I do want to add a caveat for @Systems101's sake because he was bothered by my not-too-professional "forward-looking" chatter: We can argue all day about whether VIX is itself some kind of direct predictor of future return distributions, or whether it's merely correlated due to the fact that volatility begets volatility, regardless of the existence of VIX. I suspect that the truth is really the latter, and that VIX is merely a good, convenient way to gauge where volatility, generally, is at, and thus where it's likely to go.

But for my purposes (which are not academic), I think that this is wonderful, and I don't really care if VIX is truly predictive or if it's merely coincidental, by nature, with the other factors that drive stock market volatility. All I care about is the potential for real, future volatility, and I think it's clear that VIX "predicts" that on the timeframe (a month) that I'm interested in.

talltexan

  • Handlebar Stache
  • *****
  • Posts: 1988
Re: Risk-adjusted exposure to S&P 500
« Reply #40 on: October 04, 2018, 12:34:52 PM »
Wheezle-
I didn't see this addressed in any other replies, but you'll note the Low R-squared statistic in your model: this means that the VIX explains very little of the difference in returns over time. (maximum R-2 is 1)

It seems as though your agent may have other indicators that change the allocation as well (those would improve the R-2), the more you can learn about what those are, the better.

Honestly, having someone else do this management for you is sensible for market timing. A lot of the MMM crowd here is DIY, but I think having that commitment mechanism is a good thing.

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 1375
  • Location: Seattle, WA
Re: Risk-adjusted exposure to S&P 500
« Reply #41 on: October 04, 2018, 12:36:58 PM »

What I'm saying is that if you do want to lower your portfolio volatility, adjusting according to stock volatility is a much better way since it allows you to be 100% invested the majority of the time instead of 80/20 or 60/40 or whatever you might do to lower volatility otherwise. The 10-year performance of my 50% Risk Profile portfolio is nearly identical to a 100% buy-and-hold over the period, and so it will clearly beat a "more conservative allocation," which would have been dragged down by fixed-income for a whole decade.

I understand what you are saying.  However, I'm skeptical if it actually works  over any reasonable length of time.   First, it should be stated clearly that in general people around here are buy and hold index investors, but there are lots and lots of sub-strategies in that universe, and lots of people also invest in individual stocks, and other instruments as well.    Some people are 100% stocks, but plenty of people aren't.  There is no one-size fits all.  That comment is more towards other posters than you. 

That said,  I mentioned above, it is easy to backtest a strategy for works great for 10 or 20 years, but it starts getting tough for time periods longer that.  The big real red flag in my mind though, is the S&P performance on the Safer (401)K website is wrong!  The price return for the time period was indeed 98%, but the total return including dividends was 175%!  That is one whole helluva big difference.

And if you want to tamp down some volatility you could stir in say, 30% bonds.  During the same period that would have improved the total return to 220%, and the max drawdown was only 20%.   Compare with the Safer (401)K 50% strategy claims of 134% returns with "less than 30%"  drawdown.   

From here, it looks like a lot of karate for no benefit especially if volatility is a primary concern. 


MDM

  • Walrus Stache
  • *******
  • Posts: 8775
Re: Risk-adjusted exposure to S&P 500
« Reply #42 on: October 04, 2018, 12:47:23 PM »
  The big real red flag in my mind though, is the S&P performance on the Safer (401)K website is wrong!  The price return for the time period was indeed 98%, but the total return including dividends was 175%!  That is one whole helluva big difference.
Oh dear...that puts Safer in the same league as indexed annuity sales.

See https://dqydj.com/sp-500-return-calculator/ if anyone would like to verify.

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #43 on: October 04, 2018, 01:03:30 PM »
I am emailing them right now to ask about what the exact inputs are for the risk numbers so I can act more informed about this. My colleague said it is probably the Volatility Index and the 10-year to 2-year bond spread (which is by far the most popular measure of incoming recession). I think that they should make this info available, but I don't see it on the site (it's kind of vague).

I asked about some specifics. The response:

Quote
Yes, that's correct. The only two inputs are VIX and the 10-2 Treasury yield spread, and then your own Risk Profile.

However, we believe that it is unwise to use VIX as you describe, since it partially ignores the positive bias inherent in the market as well as overstates some risks. What we're actually interested is not the VIX computation itself, but its historical "effects" on the S&P 500. So, e.g., a VIX of 12 has historically resulted in a distribution of returns with closer to 9% volatility (VIX tends to overstate real risks by incorporating a variance risk premium). This may appear troublesome because "real" VIX data can only be acquired from CBOE back to 1990, but it's made less so by the fact that VIX is not really doing anything too special, and correlates strongly to recent historical (past) volatility, much as a well-calibrated backward-looking GARCH model would. VIX could easily be replaced in this scenario -- it is only used for the sake of simplicity.

The precise distribution of historical returns at any given level of VIX is what we use to test programmatically against your 50% risk limit, so that in an enormous simulation with two million "possible universes," in no universe would you have lost 50%. This means that we're actually testing against far worse scenarios than occurred in 2008 to 2009. The true optimal allocation is that amount for which you cannot lose more than your risk limit, but for which your money grows maximally.

This is the idea behind "geometric mean optimization" as well as the more famous Kelly Criterion/Formula. Both seek to maximize the logarithm of wealth by finding that optimal allocation. In this case, however, we're maximizing the logarithm of wealth to match your bounded risk appetite. If your risk appetite "knew no bounds," you would be at the very least 100% invested, and often more, if possible (the basis for our Professional-grade data).

The 10-2 yield spread and Recession Risk is essentially laid on top of this optimal allocation, reducing it by the percentage of Recession Risk itself. So, e.g., if the optimal allocation is 70% and Recession Risk is 5%, "Your Allocation" is 65%. This means that Recession Risk ends up playing a large role only when that risk is very high, and that risk is only high when the yield curve has been flat or negative for a significant amount of time (the 10-2 spread is accurate at forecasting variance on a 3-year timeframe, so it needs to be factored in very slowly).

Regarding the performance data, yes, all of the data from 2000 is "point-in-time," meaning that there is no lookahead bias inherent in the data. If you had reallocated at market close on the last day of the month since 2000, your portfolio would have the exact same results as what we've posted (though your returns would be substantially higher owing to dividends paid -- look at the S&P 500 Total Return Index (SPXT) for a rough guide).

We hope that makes sense so far, and please don't hesitate to delve a bit deeper. We apologize for not having more detailed content just yet -- the service used to be essentially a small mailing list until only recently.

Regards,

The Safer 401(k) Team
support@safer401k.com

I may ask more about the Recession Risk metric, which they seem to have glossed over. Though tbh, I think the bulk of my questions were pretty much answered by this. I hope it answers some other questions, too (if there are details I can delve into a in a response, please let me know -- I can try to work them in).

Honestly, having someone else do this management for you is sensible for market timing. A lot of the MMM crowd here is DIY, but I think having that commitment mechanism is a good thing.
I will never, ever try to "time" the market using my gut. Having spent years trading individual stocks and options, I know my limitations... and I guess that's why I'm interested in letting go of the reins. My temperament is not suited to real "trading."

  The big real red flag in my mind though, is the S&P performance on the Safer (401)K website is wrong!
Yes, they make a point of this in the daily PDF docs. It talks about this in the performance section, that it doesn't include fees or dividends, and performance will be markedly higher over time.

@MDM, similar to your total return calculator, they're citing the "total return index." Which I assume takes into account dividend reinvestment.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #44 on: October 04, 2018, 02:33:04 PM »
You cannot logically think this ...

You can reduce volatility very simply by allocating a greater percentage to cash equivalents or fixed income.

... and this ...

Since folks here are maximizing return and accepting the volatility, I don't know why anyone here would be interested in using the VIX as you describe.
... at the same time. They are at odds.

Either @Systems101 is right that we should all be 100% invested in stock indexes and accepting all the blows all the time, or what I'm proposing is a completely legitimate way to lower portfolio volatility while maximizing stock exposure.

Sigh.  There is no way to maximize equity exposure and lower volatility at the same time.  Your proposal is neither legitimate or even founded in reality.

Systems101 never said a 100% equity position was right for everyone's risk tolerance.  Among other factors, age, i.e., time to retirement, is one of the primary considerations.  Not only is there no single equity allocation that is suitable for everyone, there isn't even an allocation that is suitable for any one individual for his entire life.

You are proposing a market timing scheme, but you insist on claiming it is something else, and you seem to take offense to anyone calling it what it is.  Hey, call it a bouncy red banana, if you want to.  It's your money, after all.  But you shouldn't be so surprised that no one else is going to call it a bouncy red banana.

Like I suggested up-thread, you should just take your little baby out on the road for a test drive, and learn for yourself.  I've got a hunch that's what you're going to do no matter what anyone else says anyway.  If your scheme goes belly up, missteps in the market can still teach you very valuable, even if expensive, lessons.

In the unlikely event that your scheme actually does outperform, the last thing you should do is plaster your method all over the internet for others to replicate.

If your scheme under performs the market, and you are comfortable with that under performance, then good for you.  Game, set, match.  Done and dusted. 

Personally, I think you are working way too hard to address a volatility concern that is trivially easy to address using more conventional methods.  But, hey, that's just my opinion.  It's worth exactly what you paid me for it.  Go ye forth, and time the market.  You have my blessing.

At this point I'm starting to doubt your grasp of the meaning of the word, "logically".
« Last Edit: October 04, 2018, 03:36:03 PM by ILikeDividends »

jacoavluha

  • Stubble
  • **
  • Posts: 232
Re: Risk-adjusted exposure to S&P 500
« Reply #45 on: October 04, 2018, 03:31:02 PM »
And where are all the actively managed funds using this strategy and outperforming?
And I mentioned this in the last thread, but again, no large fund is interested in absolute returns alone.

this is kind of my favorite part, the assertion that there's a measure that can "predict, with great accuracy, how much the S&P 500 will move" but no large fund would be interested is forming a product built around said measure

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #46 on: October 04, 2018, 03:35:24 PM »
And I mentioned this in the last thread, but again, no large fund is interested in absolute returns alone.
this is kind of my favorite part, the assertion that there's a measure that can "predict, with great accuracy, how much the S&P 500 will move" but no large fund would be interested is forming a product built around said measure
Risk-parity funds do use VIX as a measure of variance. That was the whole point of the risk-parity discussion. Tons of people use VIX for that purpose. Which is why VIX futures are so enormously popular.

The fact that "no large fund is interested in absolute returns alone" is the reason that they use VIX.

ILikeDividends

  • Bristles
  • ***
  • Posts: 374
Re: Risk-adjusted exposure to S&P 500
« Reply #47 on: October 04, 2018, 03:41:02 PM »
The VIX went over 14 today.  Do you know where your money is?



I'm a bit perturbed that the VIX didn't predict today's market swoon yesterday.  I'm waiting for my psychic to call me to explain it.  And yes, she's that good.  I usually don't even have to call her to ask.

I hope she's ok, though.  She should have called me yesterday.

On a less whimsical note, if you want to deep dive into the role the VIX futures play for an Option Market Maker's short hedge method, and why the VIX came into existence in the first place, here's a PDF you can curl up with by the fire tonight: http://www.fin.ntu.edu.tw/~conference/conference2010/proceedings/proceeding/2/2-2(A223).pdf

Too deep for me, though.  I'm not an options market maker, so it's kind of irrelevant to me.
« Last Edit: October 04, 2018, 04:53:00 PM by ILikeDividends »

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 1375
  • Location: Seattle, WA
Re: Risk-adjusted exposure to S&P 500
« Reply #48 on: October 04, 2018, 04:45:36 PM »
  The big real red flag in my mind though, is the S&P performance on the Safer (401)K website is wrong!
Yes, they make a point of this in the daily PDF docs. It talks about this in the performance section, that it doesn't include fees or dividends, and performance will be markedly higher over time.


Okay, good.  At least know their backtest is bullshit and doesn't relate to the real world.   

Raises the question why their backtest is bullshit, though.  I mean, why not run a real backtest?  Not that much harder.  Unless you want to, I don't know, conceal  under performance from the public who doesn't read the fine print. 

Also raises the question why someone would invest in a strategy based on bullshit. 

wheezle

  • 5 O'Clock Shadow
  • *
  • Posts: 94
Re: Risk-adjusted exposure to S&P 500
« Reply #49 on: October 04, 2018, 05:03:57 PM »
Raises the question why their backtest is bullshit, though.  I mean, why not run a real backtest?  Not that much harder.  Unless you want to, I don't know, conceal  under performance from the public who doesn't read the fine print. 
You're saying it's bullshit because they're making the performance numbers look lower than they actually are?

I don't know that this is something I'd complain about, especially being that dividend payouts are highly variable and completely irrelevant to the strategy.