I went last night to Kellogg’s Visiting Professor Series (despite my lack of a Kellogg degree). The official topic: Navigating Through Current Market Turbulences: A Quantitative View.
Some notes:
=-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-=
Professor Kent Daniel is the John L. and Helen Kellogg Professor of Finance. Prior to joining Kellogg, Professor Daniel taught at the University of Chicago and at the University of British Columbia.
Professor Daniel has published widely. His work has examined tests of asset pricing models, in particular tests of models attempting to explain cross-sectional predictability of asset returns and the magnitude and predictability of the equity premium. He has done both theoretical and empirical studies on psychology-based asset pricing theories. His papers have twice won the Smith-Breeden Award for the best paper published in the Journal of Finance. He is also a Research Associate of the National Bureau of Economic Research and an Associate Editor of the Journal of Finance. He received his Ph.D. in Finance from UCLA.
(His presentation will soon be available on his website.)
Kent Daniel now doing some work in this area at Goldman—he’s on sabbatical. Most economic theory based on assumptions of rationality. BUT:
There are Asset Pricing Anomalies which call those assumptions into question:
– Size effect. Small stocks disproportionately perform
o Banz 1981, Keirn 1983
– Value/Book –to-Market Effect. Stocks with lower price based on certain ratios do better going forward.
– Momentum Effect. If you buy stocks that did well over past 6-12mos., and short stocks that did poorly over last year, you get disproportionate returns.
– Issuance Effect. Firms that have issued a lot of new stock have poor performance going forward. Firms that repurchase tend to do very well.
– Accrual Effect.
These effects are very strong. The base market portfolio has a Sharpe ratio of 0.09. If you incorporate all these strategies above in an optimized portfolio, you can move the Sharpe to .461 (on a looking-back basis).
Why?
One theory: Overconfidence impacts market prices. Very clear pattern that people are overconfident. People see themselves as more able than they actually are, more able than average, more favorably than they are viewed by others. 80-90% of people say they are above-average drivers.
Individuals overweight private information –info they learn based on their own analysis. People discount unfavorable information and magnify favorable info in evaluating their own abilities.
These patterns result in overreaction, continuing overreaction, and correction phases in market pricing. i.e., bubbles/troughs.
Overconfidence predicts the patterns that we saw earlier.
People are more overconfident when they see vague information than precise information. So these biases should be stronger in stocks which require more info to evaluate:
– stocks with high levels of intangible assets (i.e., growth stocks)
– high R&D expenditure effects
– Stocks with high variance of analyst forecasts
E.g., Momentum effect and book to market effect should be stronger with stocks in this area. This in fact is what Kent and his colleagues has found.
Another study: Jiang, Lee and Zhang use as proxies for Information Uncertainty: Firm Age, Firm Return Volatility, Average Daily Turnover, and Duration of Firm’s Cash Flows (How long it takes the price of a stock to be repaid by internal cash flows).
Conclusion: the ability of Value and Momentum to forecast future returns are enhanced for high ambiguity stocks.
Overconfidence PLUS herd behavior (correlated errors) are necessary for misvaluations to occur.
=-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-=
Mansoor Sirinathsingh is a Vice President of Structured Credit Quantitative Advisory Group of JPMorgan Chase. He is responsible for developing quantitative methods for valuation of structured credit products as well as product structuring and marketing. Mr. Sirinathsingh holds a BSEcon in Finance from Wharton, as well as a BSE and MSE in Electrical Engineering from the School of Engineering and Applied Science at the University of Pennsylvania.
Credit analysts are very good at figuring out if credit rating was accurate. Where they fail is in figuring out if any particular bond was at the right price for the given rating. So Mansoor’s team worked on developing a quantitative measure of this.
Do spreads pay enough for your credit exposure?
Where along the curve is the value greatest?
Example: many high yield bonds are callable. BUT those options are only rarely called because of interest rates. Instead they’re called when the company is upgraded. These aren’t interest rate options, as some think.
New term – they analyze the “Rock-Bottom Spread”=the minimum spread investor has to be paid to get risk-adjusted return=breakeven spread + risk premium
Credit rating & Seniority & Bond cashflow pattern & Portfolio Diversity & Required return on risk ==> rock bottom spread
For every unit of risk they take on, they assume demand of a Sharpe ratio of about 0.5
What they found:
B-rated paper typically pays less than rock bottom. Investment grade pays a big liquidity premium. Using the strategy he proposes (Buy companies which are cheap according to bond price vs. fundamentals and have healthy equity returns; short companies which are expensive on a bond price vs. fundamentals basis, and have poor equity returns), he gets a Sharpe ratio of 1.2, even after high transaction costs.
In B land: rock bottom spread never exceeds actual spread. Why? High yield funds raise money by advertising a high yield. So they overinvest in the B bonds.
BB and BBB: market spread is at high spread over rock-bottom spread. Why? When bonds get downgraded to BB, there’s a lot of selling pressure. A downgrade is selling into a flooded market. Result: investors demand compensation for downgrade risk.
His recommended best trade is leveraging BBB or BB risk. CDOs are the means to do that.
=-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-==-=-=-=-=
Stephan Meili, a Director at Barclays Capital, serves as the head of model validation in the Americas and is also responsible for model validation for the credit and securitized products trading business worldwide. Previously in New York, he was responsible for credit exposure measurement at CSFB and prior to that, worked in risk management consulting at PwC. His early career was spent in quantitative asset management at UBS in Switzerland. Additionally, he has taught courses on risk management and derivatives at the Federal Reserve Bank. Stephan received a MS in Finance from Northwestern University (he spent 2 years in the Finance PhD program at Kellogg) and an undergraduate degree in Economics from University of Basel, Switzerland. He is also a Chartered Financial Analyst (CFA) and a Certified Financial Risk Manager (FRM).
If you ask nicely, Stephan may send you a copy of his complex slides on “Investing in Credit Portfolios—a Quant View”.