Author Topic: Curious about "inverse market correlation" funds  (Read 1819 times)

Reynold

  • Bristles
  • ***
  • Posts: 353
Curious about "inverse market correlation" funds
« on: September 16, 2020, 09:32:57 PM »
I'm on the financial committee of a nonprofit that is looking to invest some money for the long term, so I've discussed with them a lot of the material from these forums, since Mustachians are similarly investing for the long term.  They are generally on board with doing a 35% "cash"/65% "stocks" mix, with the stock portion being in several index funds. 

One of the other people on the committee wants, in addition to putting some of the stock market allocation into index funds, put some into inverse market correlation funds to reduce volatility.  One of the funds he wants to use is the DOG ETF, which is the inverse (1x) of the Dow-30. The other is the HIBS ETF, which is the inverse (3x) of some high-beta stocks in the S&P 500. 

I'm not familiar with either these funds or the methods they use to provide inverse returns, but I'm suspicious that they may be synthetic products that could fail if there are unexpected market moves.  I thought some folks on this forum might be familiar with these funds, and how they operate behind the scenes, and, if I'm lucky, even have data on their effects on a portfolio.  I will say the organization is willing to accept somewhat lower returns in exchange for lower volatility, but I'd rather not approve something that could fail entirely. 

reeshau

  • Magnum Stache
  • ******
  • Posts: 2508
  • Location: Houston, TX
  • Former locations: Detroit, Indianapolis, Dublin
Re: Curious about "inverse market correlation" funds
« Reply #1 on: September 17, 2020, 06:45:11 AM »
That idea is bad on multiple fronts.  Inverse ETF's are designed for *daily* targets.  Is daily your idea of long-term?

https://www.etf.com/etf-education-center/etf-basics/leveraged-and-inverse-etfs-why-2x-is-not-the-2x-you-think

Unfortunately, the suggestion would seem to be driven by fear, and in ignorance of what those ETF's are meant to do.  They are tactical tools meant for active managers.

And, returning to our standard long-term bent, if again you are investing for the long term, why would you bet against the market, when the worst rolling 15-year period since 1973 returned 6.4% per year?

https://www.thebalance.com/rolling-index-returns-4061795

Even of long-term is only 10 years, markets were up 94% of the time for rolling 10-year periods over the last 90 years.  Those are long odds to bet against.

https://www.capitalgroup.com/individual/planning/investing-fundamentals/time-not-timing-is-what-matters.html
« Last Edit: September 17, 2020, 08:26:54 AM by reeshau »

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6629
Re: Curious about "inverse market correlation" funds
« Reply #2 on: September 17, 2020, 08:17:25 AM »
Ask them to explain why the S&P 500 ETF "VOO" is -3% YTD, while 3X the S&P 500 ETF "UPRO" is -36% YTD.  Hopefully that's mind blowing enough for them to drop the 3X inverse idea.

Picture this: the S&P 500 loses 1/10th of it's value, then recovers.  What happens to a 3X ETF tracking that index?  "The answer will surprise you..." as click bait articles put it.

S&P 500:  $100 -> $90 (-10%) followed by +11.11% recovery back to 100%
3X ETF: $100 -> $70 (3x the loss) followed by +33.33% recovery back to ... 93.3% ?!

Unfortunately, that's how it works.  3X ETFs trace a path - they do not recover.  An index tracking real assets can see a recovery - but 3X funds are derivatives that simply trace a path, wherever it leads.

You could also point to 3X ETFs that shut down after massive losses (some oil ETFs), and others reducing from 3X to 2X after taking a loss.  Lose 3X and gain 2X... try making money on that!  Hopefully that's enough ammo to at least give the other person some things to research further.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6629
Re: Curious about "inverse market correlation" funds
« Reply #3 on: September 17, 2020, 08:21:09 AM »
Also, in case I come across as someone totally against 3X ETFs: I hold 3X ETF shares right now, and earlier this year I even held the exact ETF you mentioned (the 3X inverse high beta ETF).  But I did so for a brief period while predicting rising corona virus cases would damage the economy.  It didn't quite work - people ignored the rising cases and just got sick.  Even predicting the future isn't enough if you can't predict human reactions along with it.

It would be much better to stick with a combination of the total stock market and cash.  (Note this has been a volatile year for the S&P 500 - many stocks have dropped out.  You can avoid more of that by purchasing a total stock market ETF, like VTI)

Reynold

  • Bristles
  • ***
  • Posts: 353
Re: Curious about "inverse market correlation" funds
« Reply #4 on: September 17, 2020, 01:38:24 PM »
Thanks for the info, reeshau and MustacheAndaHalf, that is the kind of thing I was looking for.  The person pushing this actually wants to use it as part of a short term timing strategy, where stop-loss orders kick in on the index funds if that index drops by X%, then buys take place if it goes back up by Y%, for each fund, including these, so I think we would be swapping between these and "normal" index funds as the market fluctuates.  He has a model he has been back testing against the market to show that it produces above market returns.  Originally this model showed something like double market returns for the last 10 years, and I politely expressed my skepticism that if such a simple model worked, all of Wall Street would be using it, then he found a mistake in his spreadsheet which removed most of the excess returns, though it was still maybe 30% over the market for the decade.  He then bought access to more detailed data for the last 10 years, but that crashed his spreadsheet, so he is now writing a program to do the back testing calculations to set the X and Y percentages for optimal returns. 

Meanwhile the groups original stop-loss kicked in when the market dropped back in March, and they have sat 100% in cash (money markets and bank accounts) since then because we don't yet have a plan for where to put stop-loss orders in.  There is universal agreement that stop loss orders must be in place, as we can't afford to lose money, and I'd rather have the investments at least 65% in stocks with a mechanical stop-loss/repurchase system in place than leave almost everything in cash equivalents, as was done previously.  While I strongly suspect the stop-loss/repurchase system will cut returns somewhat compared to buy and hold, I don't think it will cut returns as much as not being in the market at all, since trading costs are pretty low these days. 

Could be worse, one person who was on the committee would have preferred something like 50% gold/50% insured bank accounts, and quit in protest at the insanity of risking any money in the stock market. 

mntnmn117

  • 5 O'Clock Shadow
  • *
  • Posts: 99
Re: Curious about "inverse market correlation" funds
« Reply #5 on: September 17, 2020, 02:16:27 PM »
I'm guessing the non-profit needs to be fiscally resilient through a major market downturn. In the non-profit world, people stop giving when the market is down?

What to do depends when the non-profit spends the money. Is there a chance the non-profit will need to liquidate most of the investments during a meltdown? In that case a high cash % is probably the best way. If the non-profit just has regular ongoing expenses then do what the post fire crowd does and be very conservative with 1-2 years of expenses, the rest in index funds giving you a 1-2 year downturn cushion.

To me a 15 year outlook and stop-losses are incompatible.

celerystalks

  • Bristles
  • ***
  • Posts: 342
Re: Curious about "inverse market correlation" funds
« Reply #6 on: September 17, 2020, 04:17:45 PM »
Thanks for the info, reeshau and MustacheAndaHalf, that is the kind of thing I was looking for.  The person pushing this actually wants to use it as part of a short term timing strategy, where stop-loss orders kick in on the index funds if that index drops by X%, then buys take place if it goes back up by Y%, for each fund, including these, so I think we would be swapping between these and "normal" index funds as the market fluctuates.  He has a model he has been back testing against the market to show that it produces above market returns.  Originally this model showed something like double market returns for the last 10 years, and I politely expressed my skepticism that if such a simple model worked, all of Wall Street would be using it, then he found a mistake in his spreadsheet which removed most of the excess returns, though it was still maybe 30% over the market for the decade.  He then bought access to more detailed data for the last 10 years, but that crashed his spreadsheet, so he is now writing a program to do the back testing calculations to set the X and Y percentages for optimal returns. 

Meanwhile the groups original stop-loss kicked in when the market dropped back in March, and they have sat 100% in cash (money markets and bank accounts) since then because we don't yet have a plan for where to put stop-loss orders in.  There is universal agreement that stop loss orders must be in place, as we can't afford to lose money, and I'd rather have the investments at least 65% in stocks with a mechanical stop-loss/repurchase system in place than leave almost everything in cash equivalents, as was done previously.  While I strongly suspect the stop-loss/repurchase system will cut returns somewhat compared to buy and hold, I don't think it will cut returns as much as not being in the market at all, since trading costs are pretty low these days. 

Could be worse, one person who was on the committee would have preferred something like 50% gold/50% insured bank accounts, and quit in protest at the insanity of risking any money in the stock market.

Since you are sitting on a non-profit, the board may be held to a prudent person standard for investing. It might be good to check with a lawyer and see what that entails in your jurisdiction. It may in essence require that the funds be professionally managed. If that is the case I think the approach would be to find the lowest cost professional manager and give the board a hard sell why it makes no sense to pay more than needed in fees. The argument will be that the funds should be managed by vanguard in a few index funds for 50-70 basis points per year rather than being managed by a high cost brokerage house at 150-175 basis points per year.

reeshau

  • Magnum Stache
  • ******
  • Posts: 2508
  • Location: Houston, TX
  • Former locations: Detroit, Indianapolis, Dublin
Re: Curious about "inverse market correlation" funds
« Reply #7 on: September 17, 2020, 06:44:21 PM »
I'm guessing the non-profit needs to be fiscally resilient through a major market downturn. In the non-profit world, people stop giving when the market is down?

What to do depends when the non-profit spends the money. Is there a chance the non-profit will need to liquidate most of the investments during a meltdown? In that case a high cash % is probably the best way. If the non-profit just has regular ongoing expenses then do what the post fire crowd does and be very conservative with 1-2 years of expenses, the rest in index funds giving you a 1-2 year downturn cushion.

To me a 15 year outlook and stop-losses are incompatible.

Both this point And @celerystalks ' point are very good.  This is not you as an individual, or a hedge fund.  You need to start with a clear investment objective in mind, and then work the mechanics around that.  It will not look like any of the board member's individual accounts.

So, something like "provide a reliable return for use in capital improvements each year" would need a high amount of fixed income investments--maybe all.  "Provide a rainy day source for years when fundraising is down" might call for some hedging, as this would be the time for withdrawals.  But in any case, the design should have some kind of expert opinion on it, because when money's at stake someone will be upset (like Mr. Gold) and may come knocking with a lawyer or bring a regulator with him.

Reynold

  • Bristles
  • ***
  • Posts: 353
Re: Curious about "inverse market correlation" funds
« Reply #8 on: September 18, 2020, 07:30:00 AM »
This is a fairly small (couple hundred member) special interest group, think a large rock collecting or astronomy club, that got a couple of decent sized bequests fairly recently, in the last few years.  Their normal expenses are covered with dues and such, they don't do any outside fund raising, so this is a fund they plan to use for special projects as desired, like some lab analysis or equipment purchases.  It would be rare that they NEED to pull any money out, I've seen once in 20 years when they had to pull out the equivalent of 8-10% of the investments for an unexpected expense.  I would consider the 35% cash portion to be plenty to cover those kinds of things. 

I was the first person in the history of the organization to formally propose an investment objective, when I gave a presentation outlining an index fund/"cash" mix like what William Pfau discusses for a long term approach.  They like it a lot, providing it never loses any significant (over 5% or so) money, ever.  Since we know that isn't plausible with a buy-and-hold approach, even with rebalancing, and a plain stop-loss system just locks in losses, I'm settling for the combo stop loss/repurchase plan since I think it will beat bank accounts, where the money is now, with a minor efficiency hit compared to buy-and-hold.   

I hope over time as they can compare the results to buy and hold, they may be comfortable enough to drop the stop-loss/repurchase complication. 

vand

  • Handlebar Stache
  • *****
  • Posts: 2300
  • Location: UK
Re: Curious about "inverse market correlation" funds
« Reply #9 on: September 18, 2020, 08:46:34 AM »
If you can't live with the volatility that a 65% stock holding brings then you should not have 65% of your portfolio in stocks... sorry if that seems a simple concept but you wouldn't believe how many people can't accept this and instead try all sorts of things to hedge, protect, insure against downside.

My go-to recommendation will always be: Start with the Permanent Portfolio. Adjust from there.

mntnmn117

  • 5 O'Clock Shadow
  • *
  • Posts: 99
Re: Curious about "inverse market correlation" funds
« Reply #10 on: September 18, 2020, 11:41:18 AM »
I was the first person in the history of the organization to formally propose an investment objective, when I gave a presentation outlining an index fund/"cash" mix like what William Pfau discusses for a long term approach.  They like it a lot, providing it never loses any significant (over 5% or so) money, ever. 

You can also do the math on acceptable losses.  Worst annual return on the stock market is something like -40%. So 5/40%=12.5% Stocks/ 87.5% Bonds. If we repeat the worst year ever then you'd see that 5% drop.

I'd vote for Permanent Portfolio, maybe get's Mr. Gold on your side and can handle a 20% drop in stocks and stay within your 5% acceptable losses.

vand

  • Handlebar Stache
  • *****
  • Posts: 2300
  • Location: UK
Re: Curious about "inverse market correlation" funds
« Reply #11 on: September 20, 2020, 04:49:19 AM »
A portfolio with a 5% max drawdown requirement is not a realistic proposition without at least having 2/3rds of it in TBills and the rest spread out in a risk parity solution.

A conservative Asset allocation is the real key here, not trading rules which turn it into an actively managed portfolio.  Having a stop loss limit on a more volatile portfolio is not any sort of solution. What will be your rules for getting back in? You will not reap the upside of a more aggressive portfolio without also being willing to tolerate the bigger drawdowns. You will just lose 5% every time the market has more than a 5% drawdown. If stop this worked then everyone would be doing it.

Reynold

  • Bristles
  • ***
  • Posts: 353
Re: Curious about "inverse market correlation" funds
« Reply #12 on: September 20, 2020, 04:18:41 PM »
Having a stop loss limit on a more volatile portfolio is not any sort of solution. What will be your rules for getting back in? You will not reap the upside of a more aggressive portfolio without also being willing to tolerate the bigger drawdowns. You will just lose 5% every time the market has more than a 5% drawdown. If stop this worked then everyone would be doing it.

That is what the person in favor of this is working on with his modeling, is to have a rule set for getting back in as well as the rule for getting out (stop loss).  He is crunching different values for the X% stop loss and Y% repurchase to see what is optimal.  I agree with the other posters here that if this approach worked, everyone would be doing it, because multimillion dollar hedge funds have a lot more resources to hire genius programmers to look for patterns.  He seems to be working towards smaller intervals for both X and Y, so I think the main effect will be to generate some unnecessary trading costs for a couple of years until it becomes obvious that buy-and-hold would work better and maybe we can switch to that.  I would definitely fight harder if ONLY stop losses were going to be used, since that would just lock in losses every time the market goes down, and he agrees that is a bad idea. 

To get back to my original post about the inverse funds, I've at least got some good questions to ask him now about those, since I was completely unfamiliar with them I wasn't even sure how to probe him on them.  Thanks for the input everyone! 

vand

  • Handlebar Stache
  • *****
  • Posts: 2300
  • Location: UK
Re: Curious about "inverse market correlation" funds
« Reply #13 on: September 21, 2020, 03:24:48 AM »
Having a stop loss limit on a more volatile portfolio is not any sort of solution. What will be your rules for getting back in? You will not reap the upside of a more aggressive portfolio without also being willing to tolerate the bigger drawdowns. You will just lose 5% every time the market has more than a 5% drawdown. If stop this worked then everyone would be doing it.

That is what the person in favor of this is working on with his modeling, is to have a rule set for getting back in as well as the rule for getting out (stop loss).  He is crunching different values for the X% stop loss and Y% repurchase to see what is optimal.  I agree with the other posters here that if this approach worked, everyone would be doing it, because multimillion dollar hedge funds have a lot more resources to hire genius programmers to look for patterns.  He seems to be working towards smaller intervals for both X and Y, so I think the main effect will be to generate some unnecessary trading costs for a couple of years until it becomes obvious that buy-and-hold would work better and maybe we can switch to that.  I would definitely fight harder if ONLY stop losses were going to be used, since that would just lock in losses every time the market goes down, and he agrees that is a bad idea. 

To get back to my original post about the inverse funds, I've at least got some good questions to ask him now about those, since I was completely unfamiliar with them I wasn't even sure how to probe him on them.  Thanks for the input everyone!


Indeed. You can have simple trend-following rules, which I believe its possible to show leads to a lower volatility portfolio  (you'll avoid larger drawdown, but you'll never plong money in at the bottom, so you won't make big gains either). However even this relies on having the discipline to implment the rules you have set yourself (so it require active management), and you could find youself getting whipsawed around.

Check out something like Meb Faber's Trinity portfolio theory for the whole gamut of global asset allocation, trend following and value tilts: https://mebfaber.com/2015/11/04/the-trinity-portfolio/
http://www.cambriainvestments.com/wp-content/uploads/2016/07/Trinity_DIGITAL_final.pdf

But again, I stress, these solutions tend to be an exercise in historic data-mining (another example is Chris Cole's Dragon portfolio), they are less likely to work as well going forward as their historic track record suggests, and max drawdown is still considerably greater than 5%.
« Last Edit: September 21, 2020, 03:34:46 AM by vand »

joleran

  • Bristles
  • ***
  • Posts: 337
Re: Curious about "inverse market correlation" funds
« Reply #14 on: September 21, 2020, 07:21:22 AM »
One interesting argument is to go long volatility explicitly as part of the non-equity holdings: https://blogs.cfainstitute.org/investor/2020/08/03/creating-anti-fragile-portfolios/