Sep

8

All Harvard and Yale would have to do to increase their endowment by 2 billion a year, would be to dollar average, putting all their money into index funds or Spiders on a once a month for 12 month basis over the next years, and eliminate 99% of their fund managers.

Russ Sears writes: 

One of the hardest things to do is to get someone to see they made a mistake because they did not have a broad enough vision or face all the the important facts, when they did have a well thought out plan but based on a narrow approach.

Kahneman's "what you see is all there is" needs to be expanded, to include the stubbornness that comes with it. This is the real danger of modeling. A good scientific model can help you overcome your emotional biases, but models are not perfect. Admitting that the models are not reality and letting yourself adapt when they are not accurate representations of what is happening is critical. This is why I believe in "counting" except when there is a liquidity crisis.

Basing performance on a Sharpe ratio has several problems / efficient frontier has several problems that it is blind to. First, it does not consider liquidity and accounting risk. Second it is exposed to modeling error on hard to model assets. And third it is exposed to execution risk or timing risk due to over-managing "exotic" assets.

A nice accounting scheme, can make certain assets almost guaranteed high sharpe ratios for the short term. Ask Gordon about how this worked with Federal Home Bank Loans and insurance companies. But same can be said about most real estate and other illiquid assets.

Second, the models assume correlations are constant and that there is no auto-correlation within the time periods. But if any asset is exposed to runs on the them, such as home ownership, then these are not valid assumptions. Structured assets are highly exposed to this risk. But so are banks and cash value insurance companies.

Finally, those that are blind to WYSIATI risk are those most susceptible to the news. Buying high because they hear how great others are doing. And then sell at the bottom because they hear how the others made a mistake. Further, most organizations that are void of valid self examination of leadership have many second tier leaders looking to say "I told you so" for any investment outside the norm, whatever that norm may be.

I have sat through many efficient frontier presentations where the conclusion was always the same; invest more in illiquid assets, invest more in assets that are impossible to model right, and invest more in exotic assets which I knew management did not have the guts to buy low and sell high.
 


Comments

Name

Email

Website

Speak your mind

1 Comment so far

  1. Andre Wallin on September 10, 2014 9:51 am

    From Welles Wilder’s New Concepts in Technical Trading Systems I created a long, short, flat data frame from indicator inputs based on closing data on the ES. From there I ran these positions against the closing prices and then I have a backtesting program in R that computes the PL. My conclusion is that the reverse of the indicators have worked the best. They have worked very well over the past three years.

    You would be best off using the trend system in a rotational market and the rotational system in a trend market.

Archives

Resources & Links

Search