I recently made a presentation of my book, The Missing Risk Premium, and thought it was concise, so I'm sharing it here.
Historical return data contradicting 'expected return positively linearly related to risk' theory:
- Within equities:
- Firm leverage
- Firm profitability
- CAPM beta
- Total Volatility
- Residual Volatility
- Financial Distress/Default metrics and equities
- Penny stocks vs. regular stocks
- IPOs vs regular stock
- Country returns in developed countries
- Country returns in emerging markets
- Analyst disagreement across stocks
- High vs. low trading volume
- R rated movies vs. G rated movies
- Volatility/future equity Index returns over time
- Overnight vs. Intraday stock returns
- Bond credit: Distress, Junk, and BBB-A rated Bonds
- Bond duration: post 2 years
- Out-of-the-money options vs. at-the-money options
- S and C corps vs. equity indexes
- Senior vs. Subordinated
- Reinsurance: rebalanced vs. peak peril
- Converts: low and high moneyness
- Merger Arbitrage: stock-financed vs. cash-financed
- Lotto vs. ‘quick pick’ lotteries
- 50-1 horses vs. 3-1 horses
- Mutual funds
- Hedge Funds
- Commodity Trading Advisors (CTAs)
- Currencies
- Futures
- Real Estate
Data consistent with
Risk/Return theory:
- Short end of yield curve
- BBB-Treasury credit spread
- Top-line equity return over libo
How high risk
generates low premium:
Winner’s curse: excess demand for volatile stocks generates
below average returns
Why the extra demand?
Why the extra demand?
- Overconfidence
- Information costs lower about risky firms
- High returns have high risk, ergo risk implies higher return fallacy
- Some are risk loving
- Some people are positive skew loving
- Alpha discovery
- Easier sell to clients (amenable to stories)
- Payoffs to fund managers
- Bounded rationality (think SML is positively sloping)
- Agency problem (exploit option with fund source)
- Those buying stocks think stocks will rise, in which case higher beta is better
Why is Flat/downward SML not arbitraged?
- Tradition (60-40 stock/equity ratio is a binding constraint)
- Irrationality (others don’t notice SML flat/negative)
- Relative risk (my favorite, other consistent data below)
- Easterlin Paradox suggests happiness is relative
- No one sells low risk, lower-than-average return stocks, because in a relative risk world, one only takes risk if the return is above average
- home bias because you want to outcompete your peers, not strangers
- Relative orientation evolutionarily robust compared to a Constant Relative Risk Aversion utility (which would be a strange coincidence)
- Glucocorticoid levels such as cortisol related to status, not wealth
- Imitation generally dominates figuring things out oneself (eg, fire, calculus), leading to an other-person directed brain
- fMRI identifies neural mechanisms for empathy, social information
- Status is a human universal, greed is not
- Reverse dominance hierarchies in humans common (ie, status more important than wealth)
- Politics about redistribution more than efficiency
Counter: If Risk has
no Premium, why take risk?
- 40% of all men reproduce, where 80% of women do.
- Men have out-of-the-money option, need to take risk
- Why not take infinite risk? Moderation in all things.
- Life is a complex, nonlinear, dynamic game where every parameter has a local maximum. Radiation, vitamin A, oxygen, tolerance, risk taking, can all be too much or too little.
11 comments:
I bought this book as an ebook from Amazon and found it very interesting.
Your conception of risk/premium confuses expectation with the real outcome. Very basic mistake.
Alas, all we have are historical data.
The thought that conditional variation in the risk premium could explain this general absence of a risk premium across this many decades and asset classes is something only a special type of theorist would say.
Your theory seems likely to be correct and so I have started to invest based on it but one one question bothers me. If the risk premium does not explain small cap's out performance of large caps what does?
I found the "40% of all men reproduce..." statement very interesting, as I've never heard that before.
However, a quick google seems to cast some doubt: http://lesswrong.com/lw/h4e/differential_reproduction_for_men_and_women/
Do you have any good references?
Baumeister mentions it here.
http://www.psy.fsu.edu/~baumeistertice/goodaboutmen.htm
johannes: you linked to a post asking for references and the most highly rated comment, with 40 upvotes, includes excerpts and citations that strongly support eric's claim. c'mon, dude!
Yes Anonymous, the link I posted does contains the 40-upvoted discussion of Baumeister's book, but also (indeed at the very top) some interesting critique of the results, which was what I had in mind:
"Apparently, it comes from studies of Y-chromosome DNA and mitochondrial DNA that indicate that the female common ancestor of all living humans is a lot older than the male common ancestor of all living humans. However, this past year, researchers discovered a man with a Y-chromosome that threw off the timeline drastically - which throws the conclusion of sex-based differential reproduction into question."
(my copy and paste lost the links to the underlying papers, but it's right there at the top of the link I posted above)
Eric: even assuming that there has been such a dramatic difference in reproduction rates in our history, do you think that is still the case today?
Jim Oliver -
The reason why small caps outperform large caps is that indexes are weighted in favor of the largest firms.
Jim: per the size effect, here's a hypothesis. I argue the equity premium is probably an illusion created by transaction costs, adverse timing, geometric averaging, peso problems, and taxes. transaction costs and adverse timing costs are even greater for small cap firms.
Eric, Amazingly I have read your books but didn't fully pick up on the fact that you think the entire equity risk premium is an illusion. I will need to re-read your last book.
Post a Comment