Do Alternative Investments Dampen Portfolio Volatility?
Devotees of alternative investments have for many years claimed that alt-heavy investors perform better than stock-and-bond investors and enjoy “volatility-dampening,” to boot.
In a recent LinkedIn post, a senior CAIA Association executive reiterated this claim, saying:
“The endowments that have allocated larger chunks to Alts materially outperform a 60/40 in the LT. More importantly, they see significantly less volatility and draw down risk.”
No hedging there on the merits of alts — more return, less risk.
It so happens that I recently examined the performance of a group of large educational endowment funds during the 10 years ended 30 June 2018. I focused on endowments with assets in excess of $1 billion that had an average allocation to alternative investments of nearly 60% over the study period. I created a composite of returns for these investors using National Association of College and University Business Officers (NACUBO) data. Then I created an equivalent-risk benchmark for the composite using returns-based analysis. (The equivalent-risk passive benchmark actually turned out to be 72% stocks and 28% bonds.)
I found that the endowment composite underperformed the equivalent-risk passive portfolio by 1.6% per year. Underperformance of 1.6% a year over a decade ain’t hay.
In the course of that work, I also examined the proposition that alts dampen portfolio volatility relative to a 60/40 portfolio. In simplest terms, I found that the annualized standard deviation of the endowment composite returns was 11.7% compared with 9.4% for the 60/40 portfolio comprising the Russell 3000 and the Bloomberg Barclays Aggregate Bond Index. In other words, the alt-heavy portfolios were 24% more volatile than “60-40.”
So much for a central element of the raison d’etre for institutional investment in alts. Over a decade, the alts-heavy endowments were more, rather than less, volatile than “60-40.”
What about performance? The diagram below is a regression of the endowment composite against the 60/40 portfolio. The slope (beta) is 1.22. The intercept of the regression (alpha) is -3.7% per year (t-statistic of -4.0).
So much for the claim that alts-heavy endowments outperform “60/40.” The endowments underperformed by a wide margin on a risk-adjusted basis, with 22% greater market-related risk.
My research reveals the much greater extent to which public market pricing is reflected in the returns of private market real estate, private equity, and hedge funds since the global financial crisis (GFC). Nowadays, alt returns are animated by returns observed in stock and bond markets.
Consequently, there is neither reason (logic) to expect alts to be “risk dampeners” nor evidence that they have been such since the GFC.
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: ©Getty Images / Baac3nes
Professional Learning for CFA Institute Members
CFA Institute members are empowered to self-determine and self-report professional learning (PL) credits earned, including content on Enterprising Investor. Members can record credits easily using their online PL tracker.
26 thoughts on “Do Alternative Investments Dampen Portfolio Volatility?”
This is a great read. Are you able to share your regression calculations for applied learning purposes?
Sure. Can you provide me an email address so that I might do so?
The regression stats are: beta = 1.22; R2 = .971; intercept = -3.69% (t of -4.0); and std error of regression of 2.11%.
My two favorite rules of investing apply.
1) Don’t try to time the market.
2)Timing is everything.
How many “average investors” are allowed to enter “alternative investments” at the bottom?
Not really surprising in a 10-year bull market where both stocks and bonds performed well. How about longer timeperiods in which (at least one) a full economic cycle is included?
“Volatility dampening” is a form of risk mitigation, no? Another form is insurance against loss from fire damage. Would you purchase a policy that contained the following language? “The insurer reserves the right to NOT honor claims for up to a decade at a time without notice.” In fy 2009, the alt-heavy endowment composite lost ~21% of its value, compared with ~14% for 60/40. For sure somebody’s house burned down in that year.
When do you need volatility dampening the most? In economic downturn. Ten years, of which most represented a bull market, don’t really provide evidence of whether any portfolio is a good volatility dampener. A longer historical period is needed to be more conclusive. Also it would be interesting to know which alts are included and in which percentages in the alt heavy 60% portfolio. Not sure if private equity for example has to be seen as vol dampener or more as return enhancer?
10 years isn’t a very long performance or risk measurement period though. What if this same exercise was done for the 1970s decade or others (if possible)?
The subject of the post is whether or not alternative investments provide reliable volatility dampening. Please see my reply to, Ingmar, above.
I agree with your overall conclusion that the benefits of alts are often overstated, particularly in marketing materials and occasionally by financial professionals that specialize in the field.
That said, I think your analysis invites some fair questions and comments:
1. Your analysis covers a single decade, which is too short a period to be considered conclusive. Have you considered extending it?
2. If you constructed an equivalent-risk traditional benchmark using returns-based analysis, then are you making the same error as promoters of alternative investments when they make dubious low-vol claims? That is, failing to properly account for skewness, higher tail-risk, financial leverage, use of derivatives, low-frequency pricing, appraisal smoothing, etc…
3. The question of greatest interest is whether reasonable allocations to alternative investments can improve a traditional, balanced and diversified portfolio’s risk-adjusted returns. I suspect the answer to this question is potentially yes, but I also believe that the optimal allocation to alternative investments is in the 0 – 40% range, i.e. much lower than the 60% average allocation that you mention above. It would interesting to see, at these lower allocation levels and with 10% position limits per strategy, whether or not alternative investments make good portfolio diversifiers.
Agreed. Have you seen Chris Cole’s work on portfolio design? He is founder of Artemis Capital and writes a fair amount. A great piece he wrote is on what he calls the “Dragon Portfolio.” He is recommending a combination of stocks, bonds, long volatility strategies, commodity trend and precious metals – all about 20% a piece.
Peter: (1) The subject of the post is whether or not alternative investments provide reliable volatility dampening. Please see my reply to, Ingmar, above. (2) I honestly don’t see a connection between the technical issues you identify (other than return smoothing) and the results presented. Nothing esoteric in what I did. As for return smoothing, I did not adjust for that; had I done so, the case for alts would be worse. See also my reply to Clayton. (3) There is a link in my post to my blog. There you will see a regression of alphas on alts exposure percentages that shows negative alpha down to 10% of assets. With alts, the more you have, the worse you did.
Richard: On point (2), to clarify: If I have understood your methodology correctly, you measured the ex-post 10-year volatility (or beta) of portfolios containing an average of ~60% alts and then you solved for the weightings of a traditional stock/bond portfolio (72/28) that would have exhibited an equal ex-post 10-year volatility (or beta) over the same period.
If that is indeed what you did, then you would have implicitly assumed that the returns of alt-heavy portfolios are lognormally distributed (they are not for all the technical reasons listed in my previous post), which is the same logical error made by alternative investment promoters when they make low-vol claims.
If the alternative investments held in the alt-heavy portfolios you measured experienced a outlier return in your already very limited sample of historical returns, then the ex-post volatility / beta you measured would be overstated, and accordingly, so would the equity content (72%) of your benchmark stock/bond portfolio.
And since the decade in question was favourable for equities, that could mean the performance differential you observed (1.6%) is either overstated or non-existent.
Peter, Can you provide evidence of “volatility dampening” (apart from that associated with return-smoothing or reduced beta) brought about by the use of alternative investments for a large sample of institutional funds? That is the subject of my post, and I dare say one of greater interest to readers of this blog than a debate of the pros and cons of normal vs. lognormal distributions in the analysis of returns.
I’ve actually been constructing investment portfolios for over 19 years, and yes, I have plenty of real world evidence that alternative investments, when intelligently combined and held in reasonable proportion, do dampen portfolio volatility – and I don’t need to data mine or back test to prove it.
Unfortunately, I also have plenty of real world examples of alts failing to provide the promised downside protection, some very recent and very extreme: in March 2020, a leveraged credit fund that was boasting a 10-year Sharpe of 3.0 experienced a 23-sigma event; that’s my whole point, volatility is a flawed and unreliable measure of alternative investment risk, and so any analysis that is based on it is also flawed and unreliable.
I acknowledge that critiquing is easy and producing original work is hard. Nonetheless, it is in all of our best interests to provide honest and fair critique even when we agree with the conclusions of our colleagues; and to accept the fair and valid critique of others, even when they challenge our work.
That was the spirit in which I posted my first comment, and also the spirit in which I post my last.
Peter, very nicely put, thank you. Best to you. Richard
This is a very interesting article: simple and factual.
Have you make a distinction between downside and upside deviation?
I would also be interested to obtain the regression details.
Please see my reply to Andri for regression stats.
I read your post with great pleasure and a smile on my face. It is so gratifying to see how meritless claims, so abundant in the social networking space today, are easily exposed with elementary analysis.
Looking forward to more of the same!
In addition to peter’s points above, you also use a US stock index, best performing region in the world the last decade. Use Acwi (or some other investable global index) like structure of most institutional portfolios and you won’t see same outperformance of simple stock bond portfolio with as much or more risk. Not defending alts per se but think this analysis doesn’t control for enough factors to condemn them either.
Agree, Clayton, that 60/40 isn’t right. When US stocks and bonds plus non-US stocks are made available to returns-based style analysis the effective equity exposure of the endowment composite is actually 72%. The standard deviation of composite and that of benchmark are identical, indicating no volatility dampening. Alt-heavy endowment composite underperforms 72/28 (w/ global equity) by 1.6% per year.
These findings reflect the prevailing incompetence of too many endowment and other institutional investment committees. As they are even unable to differentiate between outright fraud and true genius of Bernie Madoff’s and Jim Simons’ services, respectively, crap abounds of course.
In comparison, high competence obviously prevails in airlines. This results in much more reliable aircraft, particularly with regard to protecting from their most relevant risk, devastating crashes.
Thus, what is the use of focussing on annualized standard deviation instead of the worst drawdowns in our industry, where devastating crashes are also our most relevant risk? This may be the root cause of the incompetence prevailing in our industry.
However, classical trend-following Managed Futures Funds (MFFs) do make a huge difference in the alts domain. With their clear understanding of what is most relevant, they focus on exploiting the most robust and strongest momentum anomaly, driving crashes.
Their applied time-series momentum strategies provide well for the desired long volatility characteristic, negative bear beta or crisis alpha, respectively, and non-correlation to equities. All this dampens the most relevant tail risk best but also the overall volatility of equity portfolios with reliable effectiveness.
But beware of Alt. Risk Premia strategies with hidden positive bear betas. They were mixed in by some less true trend-followers to improve performance in times without strong trends after GFC. They are used even more by most other alts besides, even worse, beta from “Smart” Beta, Private Equity.
The optimal combination of MFFs with indexed equity investments as the true alpha and beta return cores is easy and cost-effective to implement and to manage. But this simple approach can provide all worth-it traditional and alternative good risk and return potentials in the markets. It reliably maximizes risk-adjusted returns based on any kind of risk significantly, making low-yielding bonds redundant.
It would be interesting to know, how many endowments use this effective approach and how this is correlated to their results in comparison to the whole sample. Has this been investigated as well or is it possible with the collected data? If yes, I would do it and post the results here if you sent me the data.
The exemplary Fortress fund is based on this unique approach, using the right alts of classical MFFs. Compared to a 70/30 world stock index/US bond index allocation it generated an absolutely outperforming return of 5,71% p.a. with significantly dampened volatility of 8,36% and max. drawdown of -17,55% since 1.2008, see: https://abrahamtrading.com/performance
Actually, this superior but simple approach is already known since 1983 from the groundbreaking “Lintner Paper”. It showed the beneficial role of MFFs as a volatile asset when combined in a portfolio with another volatile but non-correlated asset, stocks. However, this unique type of alts seems to be almost unknown. Here, it was only briefly mentioned by Ian as part of the “Dragon Portfolio”. Why so?
I find this article with compelling arguments from converted critics most helpful for a good intro: https://alphaarchitect.com/2017/10/18/trend-following-a-unique-risk-premium-explained/
The financial system limit brings with it a political system limit.
Politicians need to be aware that there are no easy answers,
no policy tools that can solve the dilemma of ‘expand now and add
to stagnation in a decade, versus do not stimulate and risk immediate stagnation,’ which is discussed in the book.
What led politicians to make our financial worse?
Economics originated as a social and political discipline but has
somehow lost its way lost in the detail of microeconomics.
The old formula of ‘stimulate your way out of recession’ worked when debt levels were much lower, nowhere near
the feasible limits, and therefore credit could expand
without anyone worrying about the consequences. As noted, for a period that
ended nearly forty years ago, that expansion caused negative real interest rates.
Now, after seventy-five years of the post-war consensus, in which
every recession has been neutered by economic stimulus, the economic cycle driven by central
banks keeps bumping up against the financial system limit.
Central banks are now the victim of their past policies.
The Fed, and other central banks, need to keep on stimulating so that more
credit can pay for the cost of borrowing resulting
from previous debt creation. The alternative is to crash the economy, which
nobody wants. The financial system limit and political system limit are interlinked.
Unfortunately, I also have plenty of real world examples of alts failing to provide the promised downside protection, some very recent and very extreme: in March 2020, a leveraged credit fund that was boasting a
Failed to establish connection to the Server
10-year Sharpe of 3.0 experienced a 23-sigma event; that’s my whole point, volatility is a flawed and unreliable measure of alternative investment risk, and so any analysis that is based on it is also flawed and unreliable.
Good article thanks for sharing this article.
nice blog and well written thanks for the proving such a beautiful article