Risks, not Risk, Again

One of the most important things I am here to teach readers is that there is no such generic concept as risk.  There are risks, and they must be handled separately.  Generic measures of risk such as standard deviation of returns, beta, etc. are unstable.  This was driven home to me when I heard a presentation from endowment investment advisors, where they talked about their models, and how the models translated current economic statistics into investment decisions.

I’m sorry, but the models can not be that good.  The financial markets are only weakly related to the real economy in the short-run, though the tie gets strong in the long-run.

Economies are unstable; get used to it.  The concept of equilibrium is also not useful, and it holds economics in thrall, because it makes the math work, even though equilibrium never occurs.

It is far better to look at your investment in an oil refiner and ask “What are the possibilities for where crack spreads will be a year from now,” than to look at the beta, correlations to anything, standard deviations, etc.  The coefficients aren’t stable.

Many advisors would rather follow a false certainty, than have to think for themselves, and have to deal with the complexity of the markets.

I am a quantitative analyst, and a very good one.  That is why I pay attention to the limitations of models, and the possibility that past data might be special, and not so relevant to the present.  It is far better that you pick over your portfolios, and ask what risks they are subject to, than to look at standardized risk measurements that describe the past or present.

Be forward looking.  What can go wrong?  Analyze each company.  Find the three most pertinent risks — read the 10-K if you are having a hard time.  See if you think the risks are worth taking.

But be assured of this.  Merely by looking at market price derived variables for stocks, you won’t learn anything valuable about the risks of what you own, or might own.  You need to think like a businessman, a sole owner, and ask whether the risks can be ably faced.

To the Consultants

Your models are garbage.  You need to review your managers at the holdings level, or you are doing no good at all.  All of the aggregate statistics hide the instability.  Far better to understand the qualitative methods of managers, and analyze whether they have a durable competitive advantage or not.

That may not seem so scientific, but science is put to bad ends in areas where there is no good science.

If I were hiring managers, I would spend a lot of time on process and people, and ignore a lot of other items.


Mathematics is of limited use in analyzing investments and investment managers.  It is far better to look for those that have good business sense, and invest with them.


  • mapkelly says:

    How do you deal with diversification? Much of the risk of holding a refinery position will be diversified away. It’s not “garbage” to try to account fire this.

  • mapkelly says:

    John Bogle says that diversification is not only the first thing an investor should think about but the second, third, fourth, and probably fifth. Large pension funds are essentially index funds with “tilts”. I can’t imagine it being worthwhile for them to think about a position in a refiner that is a small fraction of a percent of their portfolio. They’re long “the market”. Actually, they’re long a bunch of asset classes and should probably spend their times trying to get that balance right.

    • Carnegie: “Concentrate your energies, your thoughts and your capital. The wise man puts all his eggs in one basket and watches the basket.” Twain said something similar.

      I focus on industries. I can get a competitive edge there. It is desperately difficult to do asset allocation, especially using modern tools. There is no good model for it, because the correlations change radically over time. The S&P 500 and Lehman Aggregate have had both statistically significant positive and negative correlations over long time horizons. Don’t see how you can even model that risk-reward tradeoff using MPT.

      I think GMO gets it right. Calculate cash flow yields, and roughly adjust each for margin of safety. That is thinking like a businessman, rather than thinking like an academic economist, which fuels most asset allocation.

      Margin of safety, if done right, is a better tool for reducing risk than diversification. Much as I admire Bogle, I put margin of safety ahead of diversification. That might be more bumpy in the short-run but better in the long-run. Much as Buffett says that most people should use index funds, for himself, he says he would rather have a lumpy 15% than a smooth 12% return. That’s what I do and advise.

2 Trackbacks