In reading the section on factor-model based benchmarks in Evaluating Portfolio Performance, the reading mentions the normal portfolio that has exposures to the systematic risk factors that are typical for the manager. This may be a silly question but it refers to plural systematic risk factors. I guess I tend to think of systematic risk (or market risk) as THE systematic risk. What other risk factors are considered systematic risk aside from market risk?
Thanks for any help clarifying!
market risk is the largest component , but a regression would not show very high R^2 for most actively managed portfolios . Fama found that if you include premiums ( i.e. variables ) such as value ( book-to-price) and size ( small capitalization premium ) you get much higher R^2 in the regression , leading you to believe that style followed by the manager explains away most of the difference between the manager’s portfolio and market performance .
Some of the alpha left over may even be explained by a momentum variable , although no-one can quite figure out why anyone pays for it ( behavioral reasons are cited and discounted equally )
There are still other models that include leverage , currency , liquidity plus all of the above . Think of these as additional risk terms in a regression. The attempt is always to figure out how exactly the manager is loading on systematic ( i.e. replicable ) risks.
P.s. the market index contains all the risks in the above but they are diversified away . The manager, on the other hand, loads ( tilts ) on some of them i.e. weights these risks differently than the index does , so he can get a net systematic return from them in compensation for net systematic ( i.e. otherwise diversifiable ) risk
Any unexplained error in the regression is put down to the manager’s skill , which is usually tiny and usually transient.
So is any “priced-risk” an example of a systematic risk ? Again, i’ve just always equated systematic risk to mean market risk, not necessarily other risks as well.
No. If you were to buy a bond from Bernie Madoff (note: he may no longer be selling bonds), you would probably include a default risk premium; that priced risk in not systematic.
in a one-factor model , market risk ( i.e. beta ) is the only priced risk . This just means that if someone asks you why you bought the stock , you could say you expect its return to equal beta*market return , whih is fair compensation for market risk.
1-factor model doesn’t quite explain risks fully ,particularly when style bets are made , which is usually the case for active managers . So you introduce other variables that you can also call systematic risk variables because you can call them fairly priced in return for the risk specific to them .
e.g. if you buy value stocks ( even large stocks that are value stocks such as AAPL ) , then you are taking the risk that they would never go up in price for a well-understood reason ( poor prospects currently )
if you buy small stocks you bet that they won’t go bust in extreme market moves as everyone flies to safety etc.
So I’d say yes , if you buy stocks because you have a systematic reason explained mostly by experience or a regression model , then you’re taking systematic risk. one more reason could be you want to stay a close distance to the benchmark in terms of sector weights , because you’d lose your job quickly in case the overweight sectors you chose, subsequently underperformed. That’s also a systematic risk of sector movement , modeled by regressing sector returns.
Otherwise you’re trading them on basis of some information you have that is not evident to others. That would be unsystematic risk on an ex-ante basis because you couldn’t diversify it
So what makes one risk a systematic risk, and is it fair than to say that every risk falls into two buckets: either systematic or unsystematic?
Jana, so when you said “I’d say Yes”, were you referring to the question are priced risks examples of systematic risks? So interest rate risk, credit risk, liquidity risk, market risk, etc - are those some of the universe of systematic risks or is market risk a broad term for many of them?
Thank you both for the help
don’t think credit risk is systematic . Other factors portray a company as credit-risky , such as leverage or negative earnings growth or profitability , but is not amenable in a regression setup.
interest-rate could be a macro-factor and possibly could be considered systematic , but probably more evident in some industry/sectors , so could be covered more directly by loading on sector risk.
liquidity is definitely a possibility , but there is both lack of data and lack of convention on what constitutes illiquidity ( is it bid-ask , is it daily traded volume , which could have non-linear effects on your actual situation ). However researchers do consider it systematic