Financial Dogma

by

Ross M. Miller
Miller Risk Advisors
www.millerrisk.com
February 12, 2007

Much of the consulting work I do amounts to one form or another of reality checking. I got into the reality checking business while at GE, where I spent over a year of my life flitting from one GE Capital business to another in search of problems associated with the abuse of quantitative models. Reality checking is not that difficult, you look at what a model spits out and see if it makes any sense. I can say from first-hand experience that people who build models are rarely happy to receive a visit from a team of reality checkers, especially one that modeled itself after the characters in the movie Reservoir Dogs.

As a professional reality checker, academic finance journals do not make me very happy. If an article includes a model that is about companies, it's nice to see what it has to say about GE, Wal-Mart, Citigroup, Google, and any number of other idiosyncratic companies. If a model is about mutual funds, then there is a similar group of funds that one could throw at the model to put it through its paces. While lesser finance journals frequently name names in their published results, the highbrow journals rarely do and even appear to have a policy against doing so. (Who the heck is "firm391" anyway?) While there are complex psychological and sociological reasons for why this happens that I have no desire to go into here, the bottom line is that it makes it impossible to trust the results of such articles—not that anyone in their right mind does anyway. While folks at GE cowered at the thought that I might set up shop at their local Embassy Suites, shreds their models to bits, and get them fired, academics have no such worries. (I once stayed in the presidential suite at an Embassy Suites during a reality-checking visit—it was of several bizarre experiences that are reflected in Rigged.)

I do name names in my articles and I have William Sharpe to thank for it. His pioneering article on style analysis used the Fidelity Magellan Fund as an example, providing poor little me with a precedent for doing the same (and then some). It is worth noting that these results appear in the Journal of Portfolio Management, a largely practitioner-oriented journal that most top-tier finance departments do not consider as a "real journal," which means that any article published there does not count toward tenure or promotion.

Finance journals do other things that make their results essentially unauditable. It is rare than an article uses data that is readily available to anyone who do not have access to datafeeds with subscription fees that can exceed a hundred thousand dollars per year. Even then, raw data must often be cleaned before being consumed. Proprietary databases are also frequently employed. There are many considerate researchers who make proprietary data available to the masses for free, but frequently license agreements inhibit the sharing of data. Moreover, at least data provider (Morningstar—see, I name names) reserves the right to prevent the publication of any article that uses data covered by the license if they do not like what the article says. (I refused to sign the agreement, so I make minimal use of Morningstar data.)

Of course, there is a lot of finance that cannot be done without the use of expensive data, just as there are movies that cannot be made without spending gobs of money on explosions and Tom Cruise. In the movie business, however, plenty of people get by without spending lots of money, especially if they can fool frat boys and other drunk people into saying stupid things.

Long before Mr. Cohen bailed out of Morgan and Goldman and got in front of the camera, a group of Danish film directors started their own rebellion against the big Hollywood studios in 1995. They called their movement "Dogme 95," which translates into Dogma 95. This movement, which has nothing to do with the Kevin Smith flick, laid down some rules for making cheap, "realistic" movies. They required, among other things, that handheld cameras be used and that editing be kept to a minimum. Not much of anyone paid any attention to this whole Dogme 95, mostly because they were dogmatic and a little bit because not much of anyone speaks Danish, but their ideas, if not their rules, caught on. TV shows from Survivor: Ikea to Arrested Development as well as many indie movies are made in the spirit of Dogme 95.

Finance could really use something like Dogme 95, but without the rules. The underlying concept should be to create financial research whose reality is easily checked. This includes naming names and using datasets that can be readily replicated with free (over the Internet) data.

This version of financial dogma may appear restrictive, but it is not. Think about it for a moment. Suppose that you are at a top-tier finance department that spends close to a million dollars a year on data. If you not only fail to use that data, but instead use data that any schlub with a dial-up connection can get for free, you will not make friends, get tenure, or have an office any larger than a broom closet. If this isn't an arbitrage opportunity, I don't know what is.

I already have one article out that takes this new approach and it is scheduled for publication this week. I show how to compute the "true cost of active management," I name plenty of names (included dozens that are not in the article but are in an online supplement to the "Money" story about the article), and anyone with access to Yahoo! Finance or Morningstar can do the same. I have yet to hear from anyone whose reality test it flunked.

My next article, which will appear (barring a TRO) in working paper form simultaneously with my next commentary, takes things a step further. It is called "Stansky's Monster: A Critical Examination of Fidelity Magellan's 'Frankenfund.'" Using only public information from Morningstar, Yahoo! Finance, the Fama-French research database, and Fidelity's filings on the SEC's Edgar system, this article takes Magellan apart one stock at a time. And, yes, there are plenty of references to the Frankenstein movies (though not Young Frankenstein—this is a serious article); however, cinematic allusions are an optional component of the new financial dogma.