Why the Yale endowment model has potentially calamitous pitfalls according to ... Yale itself

About a week ago, Yale announced that David Swensen would be taking a temporary leave for cancer treatment. I join many in wishing him a fast and complete recovery.

We’ll have to study this article but:

- We are fans of cognitive neuroscience, brain science, not behavioral economics/finance. Our multi-year study of BE claims and research shows they do not have a peer-reviewed evidence basis, are largely a marketing/ideological campaign and do not track the best science on the brain.

- All other professions are held to evidence-based practices and claims — why not financial services? Well, business overall is clueless about brain > behavior science.

Why should financial advisors and firms be held to any different standard of proof and efficacy from say your family doctor?

The financial services industries are hyper-numbers oriented but as sales tools. Our industry pretty much ignores any independent data, evidence and proof. Again sales is always the priority. With the trillions of dollars spent and made worldwide in financial services you would think something would be spent on understanding human behavior around financial matters and the brain. Nothing is spent. Mainly our government funds this work.

The brain controls behavior — of course. The conscious mind does not, BTW.

Here’s an example, and I am listening to an MIT lecture on this as I write a client M&A report — our brains pretty much make the go/no go decision on behaviors in 150 ms instinctively and unconsciously. That’s the facts.

Not only does no one in financial services know that, most would dismiss it as useless. Stil, it is one of the physiological facts about investing behavior.

Comparatively, the behavioral economics studies are not rigorous and a lot of hand-waving and higher order concepts that are more philosophy than brain physiology.

Bottom line — we are fans of whatever knowledge and practices can predict financial behavior. Cognitive neuroscience is our best area — BE is weak — but very popular.

This could be a longer discussion but let’s start the ball rolling. Obviously, there are many things that have gone wrong.

From 30,000 feet it appears that group-think processes involving the ideologies and belief systems of around institutional investing are failing.

This can be simple or complex. For example, the misuse of the term risk when we mean uncertainty. The debunked belief in active management. The clearly abnormal expectations for alternative asset classes. We could go on. But these are all sales claims and not evidence-based. Of course, data is concocted to “prove” value — but the data is cooked, never independent and never peer-reviewed.

Getting a little closer to earth, we have seen data that diversification, for example, does not work. The use of the term irrationality is — irrational. What is more irrational — the natural ways markets behave or economic belief systems saying they must behave some “rational” way — according to econ ideology?

The equity premium is a sales claim and not a fact. Most of MPT is now Obsolete, Old Fashioned Portfolio Theory – OPT

The writer’s comments under “Behavioral Economics” are opinion and speculation. We don’t believe these statements can be made or at least need a lot of evidence to support.

Let’s remember the Nobel Prize in econ is a fake “Nobel” prize and arguments from authority are invalid. In fact pretty much nothing in econ is evidence based. Kahannamen’s work is deeply flawed but he is a great self-promoter and marketer. Unfortunately of very cartoonish ideas. The latest thinking “1” and “2” is just wrong.

“Behavioral economics had been recognized as the replacement to a failed theory.” This is simply untrue and dangerous to believe. At best, BE is a tiny little sub-area of highly speculative ideas, some of which have already proven to be damaging when put in practice.

“Information is expensive.” BE is way too cheap and easy. That’s why a lot of folks “get” it. Really, they don’t.

Boy, citing a book from 1841 is really reaching.

We’ll stop here. Frankly, the rest of the article seems speculative — at best.

Here is our point of view, which we can support with peer-reviewed research and data.

- We know very little about behavior in animals, let alone humans. – We know less about human financial behavior, although it is likely similar to “getting” behavior in other animals. – We know less about hyper-complex systems like securities markets. These are becoming increasingly complex and higher velocity via tech and globalization.

To say we know nothing is a realistic place to start. When the best and the brightest collapse an investment fund — we should hit the “pause” button, at least. We could be decades away from even preliminary knowledge about some of these things.

Now, investor’s brains and sales folk’s abhor a vacuum so some customers crave any kind of “answer” — and some sales folks crave the money they can make pretending to have answers. We have worked with portfolio managers for over 20 years and claims of skill over luck have consistently been debunked.

Anyone who claims to have any kind of answer is just wrong. Maybe we have some testable hypotheses – but realistically we are in the earliest descriptive stages.

Thanks for your insights and comments, Elmer.

I appreciate your views that financial services should be held to evidence-based practices and claims, and that they should be peer-reviewed. I also understand your view that theories are just testable hypotheses subject to needing a lot of evidence to support.

I started with the theories/hypotheses of Yale economist Robert Shiller (“From Efficient Markets to Behavioral Economics” (2003)) and MIT economist Charles Kindleberger (“Manias, Panics and Crashes: A History of Financial Crises”), developed hypotheses and tested them with daily market data and portfolios, as explained above.

I am very interested in a peer-review of the evidence that I have developed that proves the VRM strategy is effective in reducing risk due to systematic manic and panic behavior of investors which carries market prices too high and pushes them too low.

P.S. Shiller and Kindleberger do not claim credit for these theories as their own; they mention “Tulipmania” and cite historical sources, such as the 1841 reference above.

Here’s a big difference. Evidence-based means experimentally tested. Experimentally tested means: – Double blind control group testing – Replication across multiple investigators and circumstances – They must also be falseifiable – They must refer to other proven theories and findings as a explanatory basis

Bottom line – evidence-based must predict measurable events and behavior. Not backward testing, however.

Econ theories are descriptive not experimentally proven. Therefore there is no experimentally evidence that the theories and claims about things like tulipmania are impossible to prove and just story-telling.

The same is true of pretty much all of econ and certainly behavioral econ and finance.

The requirements for professional claims and evidence are not “my” preferences – they are the basic practices of all other professions and knowledge, e.g., medical, engineering, etc.

1. Ask important questions about the care of individuals, communities, or populations.
2. Acquire the best available evidence regarding the question.
3. Critically appraise the evidence for validity and applicability to the problem at hand.
4. Apply the evidence by engaging in collaborative health decision-making with the affected individual(s) and/or group(s). Appropriate decision-making integrates the context, values and preferences of the care recipient, as well as available resources, including professional expertise.
5. Assess the outcome and disseminate the results.

Translating this to an evidence-based investment decision-making process, I started with the question of whether there is an investment process that is superior to the Endowment Model of Investing, specifically with respect to reducing maximum loss, as measured by maximum drawdown from peak, while providing total returns over time that are as large (or larger).

To acquire evidence, I used a two-period, two portfolio approach. The first period was 1928-1999, and the second period was FY 2001-2011. As stated above, I used a “Walking Forward” test to the second period, which applies the investment approach to an ‘out-of-sample’ period and portfolio. This is one of the “very best methods available,” according to an independent expert. The evidence showed that the method did in fact significantly reduce downside loss while providing comparable total returns. The fourth step in the process is to apply this strategy with investors, which I am now doing.

Behavioral Economics and Finance are theories and do not provide specific investment strategies to generate evidence. A growing number of investment firms, including Fuller and Thaler Asset Management, LSV Asset Management, and Martingale Asset Management, employ investment strategies designed to exploit investors’ tendency toward psychologically induced error. What’s more, “all mean-reversion strategies, including convertible bond arbitrage, risk arbitrage, and fixed-income arbitrage, are based on the investors’ tendency to overreact.” (Source: http://post.nyssa.org/nyssa-news/2010/05/whither-efficient-markets-efficient-market-theory-and-behavioral-finance.html.)

In an IMCA conference this year, Behavioral Economist Richard Thaler stated that “mistakes in the market do happen. We classify them as either severe underreactions or severe overreactions. We then add our secret sauce and make lots of money for our clients.” (Source: http://www.advisorone.com/2012/04/23/behavioral-economist-thaler-warns-of-hindsight-bia?page=2).

FY 2012 Endowment Returns have just been reported and here is how the VRMS model return compared:

VRMS +6.94%

Yale +4.70%
Stanford +1.00%
Harvard -0.05%

Robert,

Thanks for the update. I see VRMS in the article but I’m not crystal clear on what it is?

Brooke

Brook,

“VRMS” is a Vertical Risk Management strategy applied to a constant portfolio of ten large holdings of the Exchange Traded Fund, “SPY,” so I call it “VRMS.” VRM can be applied to any portfolio. I call VRM for the energy sector, VRME.

The holdings in the VRMS portfolio are equally-weighted and rebalanced weekly.

VRM is run daily and determines what the position size will be, ranging from 0% (no positions) to 130%, where 100% is the dollar size of the portfolio.

The liquidity rule is that 25% of the positions could be added or decreased in a given day. For example, if the model wants to go to 0%, it can only sell 25% of positions on the first day.

I’d be happy to get into more details.

Thanks, Robert

Got it…so bottom line, it’s a basket of 10 big stocks, passively managed, right?

Brooke

Brooke,

Yes, a basket of 10 big stocks.

It’s considered 'active management’ since the portfolio size can change. Passive management usually refers to a constant portfolio (with rebalancing to keep the same weights).

Robert

Here’s the problem:

- Returns are data points. Like sports scores or temperatures. We don’t there freo know what caused them

- Data points without an explanation and predictions are random and as likely to 1) Be due to happenstance, 2) Be due to factors not understood

Without a fully experimentally, double-blind and peer-reviewed and replicated tested explanation of VRMS theory and applications — there is no way to claim that the theory works or the application of the theory works or is even accurate.

Without reliable predictions via the theory and applications why spend any time even talking about another theory “proven” by back testing?

Theories that don’t predict measurable data are useless. Data in the future — not in the past.

Successful models, let’s set aside the term theories, must also explain via dependent and independent variables the results they are claiming.

Looking at old data and claiming a pattern is neither theory nor model building nor knowledge. It is merely a set of abstract claims and a description. Then further claiming the hypothetical pattern is responsible for the data noted above is premature — at best.

Basic fiduciary and professionals demand such claims and statements be accompanied by far, far more care, thought and hard data and critical review.

After the recent tragic collapse of financial markets around the globe is is strikingly irresponsible and smarmy to hear a so-called intellectual and “economist” brag about “We then add our secret sauce and make lots of money for our clients.” Bernie Madoff, et all made the same claims!

Is the financial industry so gullible and desperate for ideas that a celebrity academic can get away with comparing their “professional” investing to making a cheap hamburger and people accept and quote that nonsense!? It’s reprehensible.

The Yale endowment model has outlived its usefulness in portfolio management. Institutional investors are better off pursuing shareholder activism if they seek to add value. http://alfidicapitalblog.blogspot.com/2013/11/get-rid-of-yale-endowment-model.html