Unimpressive category averages of managed futures, multialternative, market neutral, etc.

[nisiprius]

In this posting, I want to look at Morningstar's category averages for various investment categories. I don't know how to show them without displaying a specific fund first, but I don't want to look at the specific funds (in blue), I want to look at the category averages (in orange). So please ignore the blue lines. In many cases, of course, specific funds have done better than the averages--in some cases I wouldn't have heard about them if they hadn't. First, let's look at traditional securities, to see how the average large-blend stock fund and the average intermediate-term bond fund performed.

In each case: blue, throwaway fund, plotted only to get Morningstar to show me the category average; ignore it.
Orange, category average.
Green, a money market fund (virtually zero standard deviation, virtually zero correlation with stocks)

Large Blend: the average large blend fund far outperformed the money market fund. The risk was well rewarded. I'd rather have the blue line than the orange line, but the average large blend fund was a worthwhile investment.

Source


Intermediate-Term Bond: the average intermediate-term bond fund far outperformed the money market fund. The risk was well regarded. I'd rather have the blue line than the orange line, but, again, the average intermediate-term bond fund was a worthwhile investment.

Source


Now, let's look at some trendy new investment categories. Remember, I only want to look at category averages. I know that individual funds will have done better than the averages, but I am interested in the averages. This is how these categories have done in real life, real funds, real transaction costs, running real money, on the average, not just looking at the best ones.

Managed futures: a money market fund outperformed the average managed futures fund.
Source


Multialternatives: the average multialternatives fund had essentially the same performance as the money market mutual fund, winning by a nose--but at the cost of far higher risk.
Source


Commodities: a money market mutual fund outperformed the average commodities fund.
Source


Market neutral: a money market mutual fund outperformed the average market neutral fund.
Source


Low correlation? First things first. If your diversifier has less return than a money market mutual fund, it is going to drag down your return so much that it is almost impossible for any low correlation effect to offset it.

P.S. Morningstar's methodology for computing category averages.

[Rob't]

Strong support for keeping it simple, not getting tied up in complexity, and not feeling bad about it.
Thanks, Nisi!

[wickywack]

This is sobering data. I suspect all these trendy categories all had much better returns up to the point that they became trendy.

[john4546]

Someone send this memo to Swedroe.

[Bitzer]

@Nisiprius Thanks for an excellent post on a timely topic I've been seeing a lot of discussion about alternatives recently

I believe long/short alternative funds have been beating money market performance recently, as has QSPIX

@John4546...funny!

[lack_ey]

Some quick notes:

1) In a world where access to stock and bond betas is so cheap, an asset allocator considering anything else should be most interested in the active return in excess of those commoditized exposures from any prospective investment, not just the excess return above cash (though that is relevant and interesting as well). The goal would be positive active return, which even with high volatility could improve portfolios. In other words, that's positive alpha above the beta exposures, which could come from non-market factor betas, bottom-up security selection, market timing, and so on. On the other hand, even low volatility but with all returns explained by market betas on stocks and bonds would not be useful.

2) The large blend stock and intermediate-term bond categories are both terrible by this measure, as the category average simply provides almost pure negative alpha with high R^2 to the passive exposures. It's clear that you don't want to invest across the universe of available funds in these categories when the alternative is to get the beta return above cash via passive funds.

3) All the alternatives categories are hampered by even higher fees than for the stock and bond categories. You get what you don't pay for.

4) In some alternatives categories, the average fund may have significant exposures to stock and bond betas, so the results may be worse than initially thought. A fund with 0.3 equity market beta that beat cash by 1% in an environment in which stocks outgained cash by 10% a year should be considered to have had poor performance. Many multialternatives funds have net exposures to stocks, fixed income, and commodities.

5) In many alternatives categories, the difference in strategies and effectively the returns distributions are widely disparate, more so than the differences between funds in traditional categories (where all returns are strongly related to the performance of the beta exposures widely shared between them).

6) This means that the category average may be significantly less volatile than the average fund in the category. It also means that the average is not necessarily indicative of what drives the performance of individual funds. Differences in strategies and exposures are identifiable in advance. It might be more (or possibly even less) possible to pick better funds in these categories than for others.

7) All of these categories, including for large blend stocks and intermediate-term bonds, exist outside of mutual funds, and some have longer histories there. Morningstar covers mutual fund returns, which is relevant for investors choosing between mutual funds, but may not be representative of categories more broadly and what to expect.

[BigJohn]

This is sobering data. I suspect all these trendy categories all had much better returns up to the point that they became trendy.

Yup, the premise of one of my favorite books by Dr Bernstein, “Skating Where the Puck Was”.

[Portfolio7]

Nisiprius, that's a very nice piece of work! I'd done some noodling in that same territory, but with less rigor... and came away thinking that I'd best avoid the field entirely. I'll wait for some clear long term benefits before I consider anything new.

[nisiprius]

...I believe long/short alternative funds have been beating money market performance recently, as has QSPIX...

Morningstar puts QSPIX, fairly or unfairly, in the "multialternatives" category. QSPIX had its inception in 2013, so I used DVRAX in order to get a longer-term view.

QSPIX has, indeed, beaten its category average.

Generally speaking, most of what one hears about mutual funds is about funds that have beaten their category averages. People advocating one specific fund will generally give reasons why they expect it to continue to do so.

[JBTX]

In this posting, I want to look at Morningstar's category averages for various investment categories. I don't know how to show them without displaying a specific fund first, but I don't want to look at the specific funds (in blue), I want to look at the category averages (in orange). So please ignore the blue lines. In many cases, of course, specific funds have done better than the averages--in some cases I wouldn't have heard about them if they hadn't. First, let's look at traditional securities, to see how the average large-blend stock fund and the average intermediate-term bond fund performed.

In each case: blue, throwaway fund, plotted only to get Morningstar to show me the category average; ignore it.
Orange, category average.
Green, a money market fund (virtually zero standard deviation, virtually zero correlation with stocks)

Large Blend: the average large blend fund far outperformed the money market fund. The risk was well rewarded. I'd rather have the blue line than the orange line, but the average large blend fund was a worthwhile investment.

Source


Intermediate-Term Bond: the average intermediate-term bond fund far outperformed the money market fund. The risk was well regarded. I'd rather have the blue line than the orange line, but, again, the average intermediate-term bond fund was a worthwhile investment.

Source


Now, let's look at some trendy new investment categories. Remember, I only want to look at category averages. I know that individual funds will have done better than the averages, but I am interested in the averages. This is how these categories have done in real life, real funds, real transaction costs, running real money, on the average, not just looking at the best ones.

Managed futures: a money market fund outperformed the average managed futures fund.
Source


Multialternatives: the average multialternatives fund had essentially the same performance as the money market mutual fund, winning by a nose--but at the cost of far higher risk.
Source


Commodities: a money market mutual fund outperformed the average commodities fund.
Source


Market neutral: a money market mutual fund outperformed the average market neutral fund.
Source


Low correlation? First things first. If your diversifier has less return than a money market mutual fund, it is going to drag down your return so much that it is almost impossible for any low correlation effect to offset it.

P.S. Morningstar's methodology for computing category averages.

For many years I had a stake in the merger fund, a market nuetral fund that exploited the theory (and presumably inefficiency ) that acquired funds tend to go up and acquiring funds tend to go down. For many years they had respectable single digit returns with very low volatility. However in recent years it has failed to beat short term bond funds. Overcoming their near 2 percent expense ratio I guess is too difficult. I ultimately dumped it a while back. Perhaps it was an anomoly that can no longer be easily exploited.

[lack_ey]

...I believe long/short alternative funds have been beating money market performance recently, as has QSPIX...

Morningstar puts QSPIX, fairly or unfairly, in the "multialternatives" category. QSPIX had its inception in 2013, so I used DVRAX in order to get a longer-term view.

QSPIX has, indeed, beaten its category average.

Generally speaking, most of what one hears about mutual funds is about funds that have beaten their category averages. People advocating one specific fund will generally give reasons why they expect it to continue to do so.

Multialternatives is right for AQR Style Premia (QSPIX). It's also one of the most wide-ranging fund categories. After all, it's a catchall for funds running multiple alternatives strategies. Some look like regular tactical allocation funds except with a few other things sprinkled in. Others are straightforward combinations of strategies that would each fall into other Morningstar categories (e.g. equity market neutral and managed futures). It goes on.

I'm wondering what Bitzer means by "long/short alternative funds" as that's a somewhat ambiguous term as to asset exposure. Without clarification that usually means equity long/short, but sometimes it could mean in fixed income (especially credit long/short) or something else. More importantly, some people use "long/short" to mean offsetting long and short, as in market neutral. But others including Morningstar have a separate market neutral category and lump funds with substantial net positive market exposure into the long/short category. We can say that market neutral strategies are a special case of long/short that attempts to be offsetting long and short either on a dollar or beta basis.

It's trivial for the (not market neutral) long/short funds to beat cash when the market is up, as the beta exposure adds to returns and will make up for some negative alpha.

That said, even the market neutral category has beaten cash the last few years, despite all those management expenses (not to mention the cost of running the portfolios).

[SimplicityNow]

Thanks again for valuable information presented simply and with plenty of clarity.

[AlohaJoe]

Low correlation? First things first. If your diversifier has less return than a money market mutual fund, it is going to drag down your return so much that it is almost impossible for any low correlation effect to offset it.

Doesn't this apply to bonds as well? They are a diversifier with very low returns. They've returned less than a money market fund over the past 12-months and slightly more (around 0.4% more) over the past three years.

I guess I don't understand when low returns from a diversifier are a problem and when they're not.

[nedsaid]

Someone send this memo to Swedroe.

I will make a bold prediction, that is that Larry has read this and a personal message is on its way to Nisiprius.

[alpine_boglehead]

This is sobering data. I suspect all these trendy categories all had much better returns up to the point that they became trendy.

Another way to see this is that alternatives don't exist in a vacuum, they compete for capital that otherwise can go into cash, bonds, stocks etc. So if you subscribe to a universal, asset-class spanning efficient market hypotheses, prices in alternatives should also have been buoyed by low interest rates, and expected returns driven downwards.

[alpenglow]

Get with the program Nisi. These funds are guaranteed money makers...for their managers.

[nisiprius]

...2) The large blend stock and intermediate-term bond categories are both terrible by this measure, as the category average simply provides almost pure negative alpha with high R^2 to the passive exposures. It's clear that you don't want to invest across the universe of available funds in these categories when the alternative is to get the beta return above cash via passive funds...

I may not have made myself clear. I was not advocating equal-weighted investing in all of the funds in a particular category. What I was trying to do was avoid the issue of presenting one example from a universe of funds in a category, like fenestrated therbligs, and saying "this fund did great, therefore fenestrated therbligs are great" or "this fund shows that the theoretical returns from fenestrated therbligs can be realized in practice.

Looking at the average, instead of the best, is going to show you poorer returns. Since I was looking at category averages of newly-fashionable assets, I thought the fair thing to do was to put traditional assets under the same handicap by also looking at category averages.

3) All the alternatives categories are hampered by even higher fees than for the stock and bond categories. You get what you don't pay for.

That's not my fault. One has to ask "why do these funds have higher fees?" As I see it, there are two possible answers. One, which Larry Swedroe has invoked in the case of the AQR and the Stone Ridge funds, is that these asset classes are just intrinsically costly to manage. Fine, but if these asset classes are more costly to manage, then their higher cost of management is a legitimate issue with the asset class itself. It's meaningless to say "these would be great investments if you could get them at negligible cost" if, in fact, you can't do that. The other possible answer is that fund managers of the newly-fashionable assets are systematically greedier than managers of traditional assets.

4) In some alternatives categories, the average fund may have significant exposures to stock and bond betas, so the results may be worse than initially thought. A fund with 0.3 equity market beta that beat cash by 1% in an environment in which stocks outgained cash by 10% a year should be considered to have had poor performance. Many multialternatives funds have net exposures to stocks, fixed income, and commodities.

Again, it's not my problem if the supposedly "low correlations" of some asset class aren't actually low enough to provide benefit. It's the problem of the people who misrepresented the supposed "nature" of the asset class--wittingly or unwittingly.

...7) All of these categories, including for large blend stocks and intermediate-term bonds, exist outside of mutual funds, and some have longer histories there. Morningstar covers mutual fund returns, which is relevant for investors choosing between mutual funds, but may not be representative of categories more broadly and what to expect.

Sure.

[staythecourse]

Thanks Nisi for the work.

I am not sure why we are surprised. Have the different alternative guys said they expect an investor do better with them then the sp500? I don't think so. I think their point is THEIR (meaning the fund they run) expertise adds value to a diversified portfolio especially when the times get difficult for the general market (2000- 2002 and 2007-8). Even if so (many, of course, did not), the difficult aspect of active management is picking the winners out of the majority of losers BEFORE hand. There lies the problem. Unless one has a good way of finding the few winners out of the losers AHEAD of time there is no point of even considering all these different alternative approaches.

Your work showed using the benchmark clearly is not useful as a reason to add the asset class in a passive management strategy. So that means you have to figure out which ones will outperform BEFORE they are needed to do so. This is why once you are a static asset allocation guy (bogleheads) it is very hard to combine that with any active management techniques. That means if you were interested in doing alternatives you have to add MANY features of active management (which alernatives? Which specific active management alternative strategies? How much weight? When to overweight and underweight based on market conditions, etc...). When you add ALL of that active management layers it should be obvious the chances of outperforming a simple bogleheads portfolio is low (my guess under 5% over 25 year investing lifespan).

Good luck.

[SimpleGift]

It probably isn't too much of a mystery why the category-average returns of the four "alternative" asset classes examined by nisiprius in the OP have underperformed the more traditional stock & bond asset classes.

A look at their category-average expense ratios (from Morningstar):

• Managed Futures Strategies........1.89%
  Multi-alternative Funds...............1.57%
  Commodities Funds.....................1.27%
  Market Neutral Funds..................1.64%
  
  Large Blend Stock Funds..............0.89%
  Intermediate-Term Bonds............0.76%
  Prime Money Market....................0.36%

That's quite an expense headwind to overcome, if one wants to consider them in Boglehead-style portfolio.

[lack_ey]

So nisiprius, how do you interpret these results? How would you relate the category average characteristics to that of a given fund in the category?

Category average performance over time reflects the underlying performance of the average assets held minus all costs. So what does that say?

Are you interested in the averages themselves or in the behavior of funds? I definitely understand not wanting to focus on individual funds, especially those known ex-post to have been the best. But I wonder about the applicability of the averages to whatever questions you're trying to answer, which I'd like some clarification about.


What I envision as maybe a more representative analysis (but impractical for our purposes, taking too much work) would be to consider the effects of adding each individual fund to a baseline portfolio. Then we can say something like "X% of large blend funds improved portfolio outcomes, Y% made portfolios worse, Z% improved outcomes by a Sharpe ratio boost of more than 0.05 for the optimal ex-post allocation." Then compare that to funds in different categories. But something like that might not answer the questions you were looking at.

[Taylor Larimore]

Nisiprius:

Thank you for your graphs which clearly demonstrate how difficult it is to beat the market with costly alternate funds promoted by the investment industry to maximize their profits.

Morningstar: "If there's anything in the whole world of mutual funds that you can take to the bank, it's that expense ratios help you make a better decision. In every single time period and data point tested, low-cost funds beat high-cost funds."

Best wishes.
Taylor

[BigJohn]

Thanks Nisi for the work.

I am not sure why we are surprised....

Unless one has a good way of finding the few winners out of the losers AHEAD of time there is no point of even considering all these different alternative approaches.....

So that means you have to figure out which ones will outperform BEFORE they are needed to do so.

I'm betting that Nisi and many others were not surprised. The analysis shows that these alternative funds have the same issue as the active vs passive debate. IF you can pick the winner ahead of time you're OK. I'm not sure why anyone would think they can pick the winner in these categories any better than others.

[nisiprius]

What I envision as maybe a more representative analysis (but impractical for our purposes, taking too much work) would be to consider the effects of adding each individual fund to a baseline portfolio. Then we can say something like "X% of large blend funds improved portfolio outcomes, Y% made portfolios worse, Z% improved outcomes by a Sharpe ratio boost of more than 0.05 for the optimal ex-post allocation." Then compare that to funds in different categories. But something like that might not answer the questions you were looking at.

I completely agree that that something like that seems like the analysis one would like to make.

However, there's a simple-minded, and, I think valid way to look at results like the ones above, when the results that are as plain as the nose on your face. It is that takes some really extraordinary combination of high volatility and robust, persistent, reliable negative correlation, for an otherwise crappy investment (low return) to improve a portfolio with much higher return. With results like the ones above, I don't think you need to do the math. However, let's look at one. We're looking this time at the mutual fund itself, the blue line, because I can analyze mutual funds in PortfolioVisualizer, and I can't put the Morningstar averages into PortfolioVisualizer.

For example, in the managed futures category: MHFIX, Equinox MutualHedge Futures Strategy Fund.

Source

This is a fund that beat its category average. The category average would be worse. If this fund doesn't help a portfolio, the category average won't.

And if we use Vanguard LifeStrategy Moderate as a good example of a "traditional" 60/40 stocks-and-bonds portfolio, well, look at the low correlation.



So do you really believe that this fund, which earned a cumulative total of only 14% over a period of 8-1/2 years, would have improved Vanguard Lifestrategy Moderate?

Because it wouldn't have. The red curve is 80% LifeStrategy moderate, 20% MHFIX. By all means play with the percentages if you think 20% an unreasonable number.

Source



MHFIX, because it had such low return, dragged down VSMGX's return. Because it had such low correlation, it did reduce volatility. But... overall... it lowered the Sharpe and Sortino ratios.

Are you interested in the averages themselves or in the behavior of funds? I definitely understand not wanting to focus on individual funds, especially those known ex-post to have been the best. But I wonder about the applicability of the averages to whatever questions you're trying to answer, which I'd like some clarification about.

I'm interested in funds for a number of reasons. One is that as a low-sophistication retail investor, not a "qualified" investor, funds and the Investment Company Act of 1940 are suitable for me.

Another is that funds force realism. You can't hide behind cost-free academic indexes. People say, "well, of course the fund performed poorly, look at the expense ratio." I say "well, why is the expense ratio so high?" (1.70% for MHFIX!) Maybe there is a high intrinsic cost in trying to invest in this asset class.

Another is that funds force transparency. I don't know what to make of assertions that some strategy or another has been used and working magnificently for advisors or hedge funds. I just have no way to verify them. With funds, there is a huge wealth of information... and a lot of help from Morningstar.

The interesting thing is how many strategies that look great in presentations, have fallen down just about the time they were implemented as mutual funds. I don't know if that's real or just my biased perception. "Rekenthaler's Rule" is "if the bozos know about it, it doesn't work any more." Perhaps "getting mutual funds that implement them" is a good marker for "the bozos know about it."

[staythecourse]

Thanks Nisi for the work.

I am not sure why we are surprised....

Unless one has a good way of finding the few winners out of the losers AHEAD of time there is no point of even considering all these different alternative approaches.....

So that means you have to figure out which ones will outperform BEFORE they are needed to do so.

I'm betting that Nisi and many others were not surprised. The analysis shows that these alternative funds have the same issue as the active vs passive debate. IF you can pick the winner ahead of time you're OK. I'm not sure why anyone would think they can pick the winner in these categories any better than others.

Agreed. The analysis did not surprise anyone, but the take away is the inference that again for active to be successful you have to find the winners ahead of time on a time weighted basis. Active always like to talk about pick us as we are winners, but the problem is even the losers said the same thing when they first began so how does ANY investor who wants to believe in active choose between the possible available funds?

This ultimately is the BEST part of being a passive investor MORE then even the low ER. It is knowing that an index fund only job is to follow the benchmark. Stewardship alone takes care of that one.

Good luck.

[lack_ey]

So do you really believe that this fund, which earned a cumulative total of only 14% over a period of 8-1/2 years, would have improved Vanguard Lifestrategy Moderate?

Because it wouldn't have. The red curve is 80% LifeStrategy moderate, 20% MHFIX. By all means play with the percentages if you think 20% an unreasonable number.

Ex-post, I think that investing in alternatives is probably not a great idea when balanced stock/bond allocations have Sharpe ratios significantly in excess of 1. That's the window on which we're evaluating. It's hard to improve on that.

Do you expect balanced stock/bond allocations to perform similarly going forward? What would be your estimates of returns for managed futures (or some other alternatives), stocks, bonds, etc.?

Based on longer-term historical returns, current valuations, etc., I would personally say a Sharpe of about 0.33 for 60/40 seems reasonable, possibly even optimistic. That would correspond to about 3.3% above cash with 10% volatility, or something similar.

Coincidentally there's a balanced fund that had a Sharpe ratio of 0.33 since mid-2011, PIMCO All Asset All Authority (PAUIX). Let's not get too deep into the reasons why (early in betting on EM, betting against rising US stocks after valuations went high, Rob Arnott's favorite mean reversions not really happening). But just as an illustration of more ordinary returns, we find the managed futures fund improved this portfolio:

https://www.portfoliovisualizer.com/bac ... tion3_2=80

Are you interested in the averages themselves or in the behavior of funds? I definitely understand not wanting to focus on individual funds, especially those known ex-post to have been the best. But I wonder about the applicability of the averages to whatever questions you're trying to answer, which I'd like some clarification about.

I'm interested in funds for a number of reasons. One is that as a low-sophistication retail investor, not a "qualified" investor, funds and the Investment Company Act of 1940 are suitable for me.

Another is that funds force realism. You can't hide behind cost-free academic indexes. People say, "well, of course the fund performed poorly, look at the expense ratio." I say "well, why is the expense ratio so high?" (1.70% for MHFIX!) Maybe there is a high intrinsic cost in trying to invest in this asset class.

The expense ratios are high because they can find enough investors at these extortionary prices, and it seems to be working for them. Also, there's higher active share in some of these categories, more potential justification for charging high fees (the product not being as commoditized as a broad index fund). So the price wars haven't really come into effect.

Vanguard runs an equity market neutral fund with a ER of 0.25%. Some of these strats could be run fairly cheaply, maybe not as low as some index funds, but the costs need not be nearly as high.


As for the rest of your response, I was ambiguous in wording the prompt and messed up the wording, so I think you answered a different question than the one I intended. I was asking if you were interested in knowing about (1) the performance of individual mutual funds in these categories, how different funds behave, rather than (2) the performance of the average of the mutual funds. The distinction was not between mutual funds and other investment funds (e.g. hedge funds).

In some categories most funds behave similarly to each other and to the average. In others, that's not as true, even if the average fund still definitionally has the same cumulative return over the full period (the evaluation window) as the category average.

Thanks Nisi for the work.

I am not sure why we are surprised....

Unless one has a good way of finding the few winners out of the losers AHEAD of time there is no point of even considering all these different alternative approaches.....

So that means you have to figure out which ones will outperform BEFORE they are needed to do so.

I'm betting that Nisi and many others were not surprised. The analysis shows that these alternative funds have the same issue as the active vs passive debate. IF you can pick the winner ahead of time you're OK. I'm not sure why anyone would think they can pick the winner in these categories any better than others.

Agreed. The analysis did not surprise anyone, but the take away is the inference that again for active to be successful you have to find the winners ahead of time on a time weighted basis. Active always like to talk about pick us as we are winners, but the problem is even the losers said the same thing when they first began so how does ANY investor who wants to believe in active choose between the possible available funds?

This ultimately is the BEST part of being a passive investor MORE then even the low ER. It is knowing that an index fund only job is to follow the benchmark. Stewardship alone takes care of that one.

Good luck.

While identifying winners ahead of time is difficult, in some categories there are definitely differences in exposures and strategies that you could select for. Some funds are as different (within a category) as something like balanced funds and EM bond funds. If you choose a stock index fund over a bond index fund and the stock fund returns more, that's not at all on account of alpha or truly "winning" but nevertheless you could be justified in that choice.

You can also identify funds that are unlikely to provide value based on past performance and asset exposures. Gateway (GATEX, ER = 0.94%, AUM = $8.3 billion) is a US stocks + options fund that has underperformed a combination of 40% S&P 500 / 60% cash the last 10 years, with about 1% lower returns a year and higher volatility, with R^2 on market beta of 0.85 and a correlation to the market of 0.92. I don't see why anybody would expect this would improve portfolios, but apparently billions of dollars are allocated in disagreement with me.

Some alts are simply about passive exposures, and some products follow pretty basic indexes, like some of the commodity futures funds. Or some are not passive but clearly have reasons to be potentially earning money, like the TIAA Real Estate Account or something else loading on what you might think of as (non-stock and non-bond) "betas" that are available.

Other alts about trading strategy, alpha seeking (or seeking factor loadings), like managed futures or equity market neutral, could be more difficult to evaluate.

[Theoretical]

Something else about alternatives is that while you may not be able to determine in advance funds that will do well, you most certainly CAN determine that funds that are on shaky ground or are going to be terrible investments:

Examples:

1. 2%/20% hedge fund nonsense or 3% fees
2. Something with extreme manager risk, like concentrated portfolio stock picking
3: A strategy that invests in a Cayman subsidiary to invest in hedge funds (Equinox is notorious for that) rather than one that eats its own cooking (I.e. The sub is managed under the same terms as the prospectus)
4. Strategies that have low volatility and high fees or even low fees in the wrong setting. I'm thinking of Managed Futures in particular here.
5. Funds that show portfolios as just being a big swap basket and cash where the prospectus and reports make no mention of the swap fees.

Better signs:

1. Transparent strategy information and updates.
2. As with indexing, systematic or quantitative is not a cure all, but it's better than a bunch of "gut feel" investments.
3. Does it hit on more asset classes or scope, or is it very limited to one area?
4. Does the alt behave well with your portfolio? Don't expect commodities to help with a stock heavy portfolio in a recessionary situation, but they might help a bond-heavy portfolio in inflation. Similarly, QSPIX is not going to be as good a potential diversifier for a massively tilted portfolio as it would be for a vanilla 3 funder.
5. Related to 4, is it clear what kinds of markets the asset will suck in? I don't buy the notion of absolute returns all the time. That's madoffland. But you're naive if you expect a market neutral fund to blow away equities in a bull market, for commodities to do well in a deflationary situation, etc...

[Random Walker]

Had a little trouble following this, but with such diverse funds falling under the generic names “Managed Futures, market neutral, multialternative” doesn’t seem fair to group a bunch together and take average. Probably to evaluate each fund individually.

Dave

[sunnywindy]

It probably isn't too much of a mystery why the category-average returns of the four "alternative" asset classes examined by nisiprius in the OP have underperformed the more traditional stock & bond asset classes.

A look at their category-average expense ratios (from Morningstar):

• Managed Futures Strategies........1.89%
  Multi-alternative Funds...............1.57%
  Commodities Funds.....................1.27%
  Market Neutral Funds..................1.64%
  
  Large Blend Stock Funds..............0.89%
  Intermediate-Term Bonds............0.76%
  Prime Money Market....................0.36%

That's quite an expense headwind to overcome, if one wants to consider them in Boglehead-style portfolio.

Fascinating discussion. Just one point on fees: while the four alternative asset classes listed are never going to be big retail asset classes and are probably always going to command some sort of higher fee, the ETF fee compression revolution has (probably) started (to a certain point) in these asset classes. You can now buy a quasi-active broad basket commodities ETF for 0.25% ER (GraniteShares Bloomberg Commodity ETF - COMB) and an active managed futures fund for 0.59% ER (JP Morgan Managed Futures Strategy ETF - JPMF). Both these funds are very new, so it will be interesting to see how they compare in 5 years.

[BigJohn]

Had a little trouble following this, but with such diverse funds falling under the generic names “Managed Futures, market neutral, multialternative” doesn’t seem fair to group a bunch together and take average. Probably to evaluate each fund individually.

But now you're back to picking one of the few "winners" in a category in advance. Why do you think this is possible in these types of investments but not in generic active vs passive stock picking funds?

[staythecourse]

While identifying winners ahead of time is difficult, in some categories there are definitely differences in exposures and strategies that you could select for. Some funds are as different (within a category) as something like balanced funds and EM bond funds. If you choose a stock index fund over a bond index fund and the stock fund returns more, that's not at all on account of alpha or truly "winning" but nevertheless you could be justified in that choice.

You can also identify funds that are unlikely to provide value based on past performance and asset exposures. Gateway (GATEX, ER = 0.94%, AUM = $8.3 billion) is a US stocks + options fund that has underperformed a combination of 40% S&P 500 / 60% cash the last 10 years, with about 1% lower returns a year and higher volatility, with R^2 on market beta of 0.85 and a correlation to the market of 0.92. I don't see why anybody would expect this would improve portfolios, but apparently billions of dollars are allocated in disagreement with me.

Some alts are simply about passive exposures, and some products follow pretty basic indexes, like some of the commodity futures funds. Or some are not passive but clearly have reasons to be potentially earning money, like the TIAA Real Estate Account or something else loading on what you might think of as (non-stock and non-bond) "betas" that are available.

Other alts about trading strategy, alpha seeking (or seeking factor loadings), like managed futures or equity market neutral, could be more difficult to evaluate.
[/quote]

I would kindly disagree. The first part of your post is reasonable as a critique on Nisi's methodology of his analysis stating he did not choose appropriate benchmarks for each alt approach. I have no interest, but if you do claim that then it is your responsibility to redo the analysis based on using the correct benchmarks. Unless you or someone else does then his analysis stands. BTW I would not be surprised if you are correct, but need the data to support it instead of just throwing stones on his analysis.

Second, if one fund does well OR not has no implications on their performance going forward. That is the whole point of the SEC label of "Past performance..." after the active fund marketing following he glorious late 90's in the terrible early 00's. Unless somehow alt funds are different EVERY study has shown the same as Jensen's study in 1930's that today's good fund is tomorrows bad performer. So you can't say, If one is good today then put money in this one as it will do well tomorrow."

Good luck.

[Random Walker]

Had a little trouble following this, but with such diverse funds falling under the generic names “Managed Futures, market neutral, multialternative” doesn’t seem fair to group a bunch together and take average. Probably to evaluate each fund individually.

But now you're back to picking one of the few "winners" in a category in advance. Why do you think this is possible in these types of investments but not in generic active vs passive stock picking funds?

Well, not back to picking winners, but focus on just a specific passive formulaic example. For example, I have QSPRX market neutral style premia fund. it invests in 4 styles across multiple asset classes. The 4 styles are value, CS Momentum, defensive, carry. The asset classes are stocks, bonds, commodities, currencies, interest rate futures. It is market neutral. How many other funds are comparable? Not rhetorical question, I really don’t know the answer, but my guess is very few. But essentially meaningless I think to compare this to anything but something at least using same styles and asset classes, same proportions, same leverage, and likely a bunch of variables I don’t know about.
For those interested in these type of funds, I think one needs to use the following logic. Use Larry’s criteria for factors: persistent, pervasive, robust, intuitive, implementable. If the factors pass these criteria for the individual, then put your trust in a firm known to be a big, efficient, reputable player. The backward looking track record of these funds is likely too short to be meaningful.
Another example, Managed Futures is sometimes synonymous with commodities. But some managed futures funds access time series momentum through multiple asset classes, not just commodities.

Dave

[lack_ey]

I would kindly disagree. The first part of your post is reasonable as a critique on Nisi's methodology of his analysis stating he did not choose appropriate benchmarks for each alt approach. I have no interest, but if you do claim that then it is your responsibility to redo the analysis based on using the correct benchmarks. Unless you or someone else does then his analysis stands. BTW I would not be surprised if you are correct, but need the data to support it instead of just throwing stones on his analysis.

Second, if one fund does well OR not has no implications on their performance going forward. That is the whole point of the SEC label of "Past performance..." after the active fund marketing following he glorious late 90's in the terrible early 00's. Unless somehow alt funds are different EVERY study has shown the same as Jensen's study in 1930's that today's good fund is tomorrows bad performer. So you can't say, If one is good today then put money in this one as it will do well tomorrow."

Good luck.

From your response, I'm not sure if I successfully communicated anything at all, as I didn't even criticize nisiprius's methodology of the analysis there. There's not two different halves.

Let me put it this way: true alpha is inconsistent and largely driven by luck in the vast majority of cases. Because of the randomness, it's difficult to pick between funds doing similar things. Past winners will not tell you about future winners. However, it is possible to observe and understand more about asset exposure and strategy. For example, I can tell you that a total stock market index fund and a low volatility index fund are going to be a little different, with the low vol fund likely to have lower vol going forward. Short term performance may favor either. Conditioned on a bull market, the total market fund is likely to have higher returns. The factor research out there suggests higher Sharpe ratio for the low volatility stocks over the long run, though perhaps with higher correlation to bonds. All this is based on asset exposures and risks taken.

In some alt categories there are large differences in asset exposures and risks. We can identify based on stated fund strategies, holdings, and past performance the types of properties to expect, at least to some degree. This can tell us a little bit about forward return in the same sense as the previous example with total stock and low volatility stocks.

One of my hypotheses would be that expensive funds that primarily repackage stock and bond risks, perhaps with some discretionary market timing thrown in, are likely to have significant correlation with stock and bond portfolios while being unlikely to add much of value, especially if there's no good story behind whatever the active share is coming from. I don't know which of these funds will do better, but I would avoid everything that looks like this (in favor of funds pursuing other risks).

One example would be something like IQ Hedge Multi-Strategy Tracker ETF (QAI, ER = 0.76%) in the multialternatives category, a $1.1 billion ETF that I saw an article on in ETF.com a while back. It holds other ETFs, primarily well-known stock and bond ETFs, trying to replicate hedge fund beta to some extent. Check the factor exposures:
https://www.portfoliovisualizer.com/fac ... sion=false

If an equity market neutral fund is intentionally and specifically loading on equity factors, that can be seen in the returns, and we could allocate based on that and other known attributes. Non-market factors may well have some predictive power to identify long-term winners, even if any given factor has a very high chance of not providing a positive return in any given year.

[cheapskate]

Look at the bright side, these products go down regardless of what the equity and bond market, providing you with a perennial source of Tax Losses to harvest. Trust me, I own some AQMIX/QMHIX (AQR dog poop that I picked up several years ago, before the S&P went up another 50% on me). These have been a source of Tax Loss joy for me since I bought 'em - I just switch between the 2 several times a year. Sadly that will stop once they hit 0.

Also, if investors shun products like these, how will Cliff Asness get paid to produce his pompous, intellectual-sounding white papers that provide so much fodder for discussions here ? The forum would get pretty boring...

[wickywack]

In some alt categories there are large differences in asset exposures and risks. We can identify based on stated fund strategies, holdings, and past performance the types of properties to expect, at least to some degree. This can tell us a little bit about forward return in the same sense as the previous example with total stock and low volatility stocks.

Nisi's analysis seems to call this into question. All of these various investment categories were presumably created because someone expected them to do well based on strategies, holdings, and past performance.

In some cases, we're now seeing nearly two decades of trailing a money market fund. To me, that's a troubling pattern: these relatively new investment categories are created based on data that suggests good forward returns and low correlation to equities. Once they're established, that fails to actually hold.

[lack_ey]

In some alt categories there are large differences in asset exposures and risks. We can identify based on stated fund strategies, holdings, and past performance the types of properties to expect, at least to some degree. This can tell us a little bit about forward return in the same sense as the previous example with total stock and low volatility stocks.

Nisi's analysis seems to call this into question. All of these various investment categories were presumably created because someone expected them to do well based on strategies, holdings, and past performance.

In some cases, we're now seeing nearly two decades of trailing a money market fund. To me, that's a troubling pattern: these relatively new investment categories are created based on data that suggests good forward returns and low correlation to equities. Once they're established, that fails to actually hold.

Sorry, but I don't understand how that follows.

Inidivudal funds are created, not categories. Morningstar groups funds into categories. I'm saying that different funds within a category can have identifiably different strategies and asset exposures. We can understand to an extent what will drive their returns going forward.

Some fund strategies will not have low correlation to equities. Others by the construction of the underlying funds will and have, such as equity market neutral. In some of the categories that do not, we can figure out which funds will not have low correlation to equities and which are likelier to have lower. Fairly trivially, funds that on average have net long stock holdings are probably going to be correlated.

Relative returns are extremely random and more difficult to predict, but to some extent, higher levels of probably compensated risk may provide a higher chance of greater forward returns. Those loading on factors or well known, perhaps persistent (over the long run) strategies that are not getting milked dry could have positive returns above cash. That's a similar argument as to those saying that maybe small value stocks will outperform the market, with more risk, over the long run. We can identify small value stock funds from broad market stock funds in advance. It's not difficult or controversial.

Certainly alternative investments based on more beta-like exposures to risky assets should have some justification for having positive returns. e.g. direct real estate, P2P lending, short volatility. They could provide positive active return in excess of cash and the regular stock/bond exposures. Many alts on the other hand are based on trading strategies and are probably more speculative, though some of these truly may be hedged and have basically zero correlation with traditional assets.

I don't think all the asset managers create funds because they think they will do well. They want to gather assets and charge high fees. Doing well helps but that's hard.

[whodidntante]

I've been pleased with AQR Market Neutral QMNNX (poor man's share class). I was hoping that Vanguard would get their act together for their own market neutral fund since it's less expensive, but it seems they have conflated "market neutral" with "gain neutral." At least, so far.

[nisiprius]

Looking at individual funds is like the people who say--with a straight face, I think--that active funds are superior to index funds and that the SPIVA results are misleading--because investors are smart enough to invest in the good active funds, not the bad ones. Why, just look--all the active funds in the 401(k) plan have five stars, while the index funds only have three.

[lack_ey]

Looking at individual funds is like the people who say--with a straight face, I think--that active funds are superior to index funds and that the SPIVA results are misleading--because investors are smart enough to invest in the good active funds, not the bad ones. Why, just look--all the active funds in the 401(k) plan have five stars, while the index funds only have three.

Are you talking about examining individual fund characteristics, risks, asset exposures, etc. or looking at relative returns? True alpha is highly variable and in most cases probably just primarily driven by noise/randomness, being unpredictable.

But there are identifiable differences between funds.

If we grouped all the funds in a 401(k) plan as a category, including different US stock, international stock, and various bond funds, I would suggest that this category average doesn't have much bearing on the characteristics of any given fund in this category. And while it's very hard to predict which will perform the best in the next year, in the long run I would guess the stock funds are more likely to return more. For some alt categories (well, multialternatives at least) the differences between funds are on this level, and the category average about as meaningful, though it may be more difficult to predict relative long-term returns. In some others differences can be substantial but maybe not as great.


SPIVA provides stats across the range of funds. That is useful but it doesn't in of itself prove that it's impossible to select superior funds. We know that selecting based on past relative performance alone is not a winning strategy (though many follow this approach, wrongly). Selecting funds randomly or choosing all of them also doesn't work. In fact, there are a lot of possible losing strategies. But that doesn't mean there aren't some indicators and information that could be used successfully on average for some modest edge at least. Maybe there isn't anything ex-ante but I wouldn't be too confident in a judgment either way.

It's widely accepted and known that selecting active funds based on expense ratio improves results relative to the average (though not enough to outperform the index), with the causal mechanism being clear. There are a number of other measures that could be used in conjunction with that. One that was touted years back was active share, though closer inspection reveals that just increases the dispersion of alpha and not the mean, which kind of makes sense. In any case, some of these may legitimately be predictive, perhaps by enough.

[Taylor Larimore]

Nisiprius:

Thank you for your important and timely opening post (OP) and your excellent analysis!

Best wishes.
Taylor

[BigJohn]

Well, not back to picking winners, but focus on just a specific passive formulaic example. For example, I have QSPRX market neutral style premia fund. it invests in 4 styles across multiple asset classes. The 4 styles are value, CS Momentum, defensive, carry. The asset classes are stocks, bonds, commodities, currencies, interest rate futures. It is market neutral. How many other funds are comparable? Not rhetorical question, I really don’t know the answer, but my guess is very few.

Dave, I understand your point and don't know enough about the alternative investments to make any different/better comparisons than Nisi did. However, from your comments above you are at least having to pick a winning premia formula on the assumption that fund A's formula is a winner and fund B's formula is a loser. Maybe it's just my ignorance about premia (full disclosure...I'm not really an advocate) but this still seems like an exercise in picking winners and losers.

[Random Walker]

BigJohn,
I agree picking winners based on performance not the greatest idea. That’s why I said pick the premia you believe in, and choose the fund(s) that faithfully implement them in the most true, passive, cost efficient manner you can find. Sort of look forward more than backward. I can see why people may avoid these with such short track records. I think for those potentially interested in investing, I’d look for faithful execution of the strategy more than how the strategy has worked out over the short recent past.

Dave

[HomerJ]

That’s why I said pick the premia you believe in, and choose the fund(s) that faithfully implement them in the most true, passive, cost efficient manner you can find.

What premia do I believe in? How do I pick from the list of possible premia?

How do I know which fund will faithfully implement them? Are these funds that transparent? Can I see all their trades? How much research do I have to do before I can understand each trade, and accurately calculate how faithful these funds are in implementing the premia I decide to believe in?

What if I pick the wrong premia to believe in?

[james22]

This is sobering data.

Do you find it sobering to realize you're insurance costs without a claim?

I suspect all these trendy categories all had much better returns up to the point that they became trendy.

I suspect betting on beta and buy-and-holding without regard to valuation will never be more trendy than at the market's top (having had much better returns up to that point).

[wickywack]

This is sobering data.

Do you find it sobering to realize you're insurance costs without a claim?

Sorry - not following you here.

I suspect all these trendy categories all had much better returns up to the point that they became trendy.

I suspect betting on beta and buy-and-holding without regard to valuation will never be more trendy than at the market's top (having had much better returns up to that point).

Perhaps, though there are certainly no shortage of threads or articles on either a coming correction or low expected returns on equities going forward. I think a difference with stocks and bonds is that they've been around for centuries and folks have at least some intuitive understanding of their underlying valuation.

In contrast, I don't have an intuitive understanding of what makes, e.g., a market neutral fund go up or down over the long haul.

[aristotelian]

Appreciate these charts, as always. I would like to see the results after the next bear market. It's not surprising the hedging strategies have underperformed since the event they were designed to hedge against hasn't happened. I suppose the level of underperformance is what is most remarkable here.

I would also be curious to see what the numbers would be if the hedging strategies were compared without considering fees. VMNFX has an expense ration of 1.6%. That has got to be a huge drag on the returns. How much are these results an indictment of the fees versus the strategy itself?

[james22]

Insurance is a hedge (with costs), wickywack. One isn't usually unhappy that (as aristotelian answered) the events it was designed to hedge against didn't happen.

And I don't believe it possible that many have an intuitive (or otherwise) understanding of the Fed policy on valuation.

[Random Walker]

That’s why I said pick the premia you believe in, and choose the fund(s) that faithfully implement them in the most true, passive, cost efficient manner you can find.

What premia do I believe in? How do I pick from the list of possible premia?

How do I know which fund will faithfully implement them? Are these funds that transparent? Can I see all their trades? How much research do I have to do before I can understand each trade, and accurately calculate how faithful these funds are in implementing the premia I decide to believe in?

What if I pick the wrong premia to believe in?

All good points and questions HomerJ. To choose the factors to believe in, I’d start with Larry Swedroe’s factor book. He explains the criteria to use for weeding out the 5 or 6 significant factors from the 600 or so that make up the “Factor Zoo”. The criteria are persistent, pervasive, robust, intuitive, investable. Each of those words has a specific, meaningful, and useful definition. As far as choosing specific investment vehicles, the book makes recommendations at the end. I admit, choosing the vehicles is beyond me, and I use an advisor for that. I use funds recommended in the book.

With regard to the question “what if I pick the wrong premia to believe in?”, that is perhaps the strongest reason to diversify across premia. Market beta could potentially be the “wrong one”. If you diversify broadly across sources of risk/return, you will still have plenty of market beta in your portfolio.

Dave

[lack_ey]

I would also be curious to see what the numbers would be if the hedging strategies were compared without considering fees. VMNFX has an expense ration of 1.6%. That has got to be a huge drag on the returns. How much are these results an indictment of the fees versus the strategy itself?

That figure includes borrowing and dividends costs paid on the short side, which is IMHO an incredibly stupid way of looking at things. That's a cost of implementing the strategy and is offset by dividends earned on the long side. The expense ratio excluding those things (so the management fee and other miscellaneous expenses) is 0.25% for the Vanguard fund. That's not been the problem.

Note though that Vanguard only got the fund in the mid 2000s (the older history is before it was under Vanguard), and it didn't get rid of non-Vanguard subadvisors until 2010.

For a lot of the funds, the actual fees are in excess of 1% and are a big problem, just not the biggest problem. It depends on the category and exposures, really. I mean, the commodity funds have done terribly mostly on account of commodities futures being terrible the last 10 years (unexpected deflationary shock, lower inflation overall than expected, slowdown of demand especially as China's economy changed, fracking / technological / other advances, heavy contango in some futures through the period on average).

For funds relying on trading strategies, relative returns rather than asset betas to generate excess returns above cash, that's never going to be consistent and expectations should be low generally. I don't know what kind of returns people expect.

[siamond]

I never bothered to run such numbers myself, shaking my head at the dubious premise underlying those 'alternatives', and quickly moving on to topics I deem more interesting. I didn't expect the outcome to be that scathing though. Thanks, Nisi, for bringing such hard facts to the table.

The more I look at historical data, the less I believe in the idea of *reliable* asset class diversifiers (besides the basic bond/stock split). It now seems to me much more important to learn to live with volatility, instead of trying to fight it teeth and nails. There are much more significant risks to consider when looking at a full retirement period, imho.

[Random Walker]

I never bothered to run such numbers myself, shaking my head at the dubious premise underlying those 'alternatives', and quickly moving on to topics I deem more interesting.


Siamomd,
I intuitively super strongly believe in ruthless market efficiency too. This is what makes the behavioral anomalies so fascinating. Markets have been efficient for a long time and are only getting more efficient, yet despite gobs of information, computers, math, gizmos, and people competing against one another, these behavioral anomalies seem to persist. A behavioral component to value, the small growth lottery effect black hole, the poor performance of IPOs, cross sectional momentum, time series momentum are especially interesting because of their potential “dubiousness”, yet they seem to persist.

Dave