Predictive or Market Based Investing?

As I mentioned in my previous post, the last piece of the puzzle when constructing your portfolio is selecting a management strategy.  There are two broad categories of investment style that I will discuss; predictive and market based.

A predictive strategy aims to outperform a given benchmark by selecting securities that the fund manager feels are undervalued, or are experiencing momentum in their price (depending on whether it is a value or a growth fund).  This management expertise does come with a price, and as I have previously mentioned, it is very difficult to determine which securities will outperform.  Excellent active managers do exist, and they can produce impressive results.  

Noah Blackstein's fund, American Power Growth (Dynamic Funds), has produced a compound annual growth rate of 9.3% over the past 10 years; Blackstein is clearly doing something right.  Now that we know about this fund shouldn't we buy into it?  In my opinion, buying into this fund now would be like buying into Apple at $700/share.  By the time you have noticed that a fund or a security is outperforming, it is probably too late to capture any gains; it can be argued that a strong manager is likely to continue to outperform, but historical data shows us that the probability of a predictive fund outperforming its benchmark decreases significantly every year.  A graphical representation of actively managed Canadian Equity funds that outperform their benchmarks can be seen below.  The blue bars represent the percentage of all actively managed funds that beat their benchmark in one year, two consecutive years, and then three consecutive years, and the orange bars represent the number of the original sample of funds that remain after each year.  So after three years, only 2.7% of actively managed funds beat their benchmark every year, and only 68.5% are still in existence.


Knowing this information, I believe that picking managers is the same kind of gamble as picking stocks is.  There is only so much money in the financial markets, so if one fund is performing very well, there is a fund somewhere else performing poorly to allow the capital to shift.  All the while, the investors in these actively managed funds are paying a premium to the managers.  In other words, it's a zero-sum game for investors, and a negative sum game net of fees.

By contrast, index funds don't compete with active-fund managers; they simply follow an entire index and save mightily on fees and trading costs. In a 2010 paper with frequent collaborator Kenneth French of Dartmouth, Fama found that around 65% of the more than 3,100 mutual funds studied underperformed passive portfolios of similar risk in the long run. "And the investors who generate big returns over five years, the guys they write books about, are supposed to keep winning, right?" says Fama. "Well, they don't."

A market-based investment strategy, also known as indexing, is the idea of capturing the entire universe of stocks contained in a given index in order to track the performance of that index.  Because there are minimal research costs, and trading costs are low, passive strategies have significantly lower overall fees than predictive funds.  If it were the case that active (predictive) management had a strong track record of beating their benchmark indices, then they would make a lot of sense, but this is not the case. The costs of research and trading in an actively managed fund are expensive, and the return on the investment has not been historically beneficial.  If actively managed (predictive) funds are charging high fees and not beating their benchmarks, why not just buy the whole benchmark?  

Further to the idea of indexing, using academic research in economics, financial markets, and statistics, it is possible to adjust an index portfolio to track specific risk factors that have been shown to increase expected returns while still maintaining low costs.  As their name suggests, the risk factors do increase risk, but because the risk is known and calculated, it becomes possible to construct a portfolio that will diversify against these risks and result in higher expected returns and lower overall volatility.  The low costs in these portfolios are derived from the fact that individual securities become interchangeable within their asset classes and geographies which reduces the volume and frequency of trades required to maintain a desired asset mix.  This method of investing eliminates the need for exhaustive research on any given security and shows that using science to invest can result in robust, low cost portfolios.

Investors will have different preferences on the way that they want to invest.  If you feel that it is best to try and outperform the benchmark with an active strategy, it is important to spend time selecting a manager; the right manager will consistently exercise logical and rational strategies that fit into the ideal asset allocation and level of diversification that you have established.  It is my opinion that the time and effort spent searching for the right manager is time wasted when low cost engineered portfolios constructed with an academic approach are available - remember, it's not timing the market, it's time in the market.  I believe that a passive portfolio that captures the global universe of stocks with a carefully constructed diversification strategy and a factor weighted tilt is the ideal way to invest.  If you choose to take a risk in the zero sum game of playing the markets, just how much will the increased fees affect your long term performance?  A lot.  This is a $1,000,000 dollar portfolio over 30 years...the difference a percent makes over that time period is drastic.