CN's Blog

I am interested in doing a research topic in finance.  The current topic I am exploring is Adaptive Markets Hypothesis.  The idea that focusing on how markets evolve over time appeals to me more so than assuming that humans are rational and that information would be distributed freely (Efficient Markets Hypothesis).  I learned in my institutional economics course that institutions tend to be path-dependent, meaning that their past states influence what kind of equilibrium they are able to reach.  Financial institutions are no different, and therefore one must look at the changes over time to find what investment strategies would be profitable.  Here are a few interesting conclusions from the Andrew Lo paper on AMH:

• It is possible to make excess profits in financial markets without taking on excess risk
• Risk/reward for any given strategy will change over time as other investors catch on
• Sticking to the same investment strategy will not produce the best returns (it may in the short run, but not in the long run)
• Innovation is the key to excess returns

Another interesting puzzle was the Grossman-Stiglitz Paradox, which states:

If a market were informationally efficient, i.e., all relevant information is reflected in market prices, then no single agent would have sufficient incentive to acquire the information on which prices are based.

So that basically covers what I am preparing to write about this year.  Some other potential topics I might go with include liquidity crisis or abnormal hedge fund returns.  I strongly encourage you to leave some kind of comment or feedback.  Thanks in advance.

A few books on the top of my reading list:

Investing: The Last Liberal Art by Robert Hagstrom

Advances in Behavioral Finance by Richard H. Thaler

Market Wizards by Jack D. Schwager

We talked very briefly today.  Prof. Rivkin thought the ideas behind Andrew Lo's Adaptive Markets Hypothesis were very interesting, but he was not convinced it was a better alternative over EMH.  For next week, I need to have the following sections of my thesis written:

• Framework
• Data Set (characteristics)
• Write up of Autocorrelation and Geary Runs tests

In our meeting today, we briefly talked about the thesis presentations last week.  I plan to e-mail Professor Ishii and respond to his concern about the long-term differenes between EMH and AMH (my response is basially that there aren't any, but the issue I am trying to address really is in the transition phase, in which the two views do differ).  Prof. Nicholson's concern about the significance of the filter trading rules is more of a direct problem that I need to address.

What I will accomplish over Interterm:

• Run all of the necessary autocorrelation tests
• Develop a criteria for significance of filter trading rules
• Look at points in history in which efficiency might be affected in the Hong Kong stock market

In our meeting today, we discussed autocorrelation testing.  I had started redoing the Stata project using another built-in ACF test.  Ironically, the autocorrelation test results were excessively conclusive about the evidence of autocorrelation in our data.  Since this looks too good to be true,  will spend this week figuring out exactly what I did wrong to get such impossibly good results.

Also, I was told that there's a prominent Amherst alum working at a well known asset management company that would be interested in talking to me about my thesis idea.  I take this opportunity to welcome anyone reading this to contact me even if it is nothing more than just letting  me know you are interested in reading about this topic.  Or that your heart skips a beat every time you see me publish a new post.  Feel free to leave comments or send me an e-mail.  I'm tired of seeing 4-5 new spambot trackbacks on my blog everyday and not much else.  There is nothing more valuable to me than hearing your opinions.  Thank you.