Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion

Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.

    Support MFO

  • Donate through PayPal

Mapping Investor Odds-Part 1

MJG
edited April 2013 in Fund Discussions
Hi Guys,

Kenny Rodgers said it best of all in his song “The Gambler”. He sang: “you got to know when to hold’em, know when to fold’em, Know when to walk away, and know when to run.” That’s a pile of choices. A superior decision depends on assessing each event’s odds and its payoff matrix.

The 911 terrorist attack is a tragedy that directly caused the deaths of 3000 folks. That incident prompted some of us to substitute auto travel while abandoning air travel. Statistically, on a per passenger mile traveled basis, the auto death base rate is far more hazardous than the same metric is for air transportation. The statisticians estimate that an additional 3000 secondary deaths can be attributed to the unfortunate choice in the alternate travel mode. Many choices, some wise, some not so prudent.

Knowing the odds is a primary consideration in any gambling adventure, and odds are silently rooted in most of life’s decisions. Most folks would acknowledge that it is necessary know-how when applied to investment decisions. Recognizing the odds is directly coupled into knowing and understanding statistical data.

As Gary Belsky and Thomas Gilovich observed in their “Why Smart People Make Big Money Mistakes” book: “Odds are you don’t know what the odds are.” The purpose of this posting is to eliminate that deficiency by defining some of those odds when making active/passive mutual fund management choices.

Ed Thorp gained fame for his analysis and development of Blackjack odds tables to maximize payoffs with proper play decisions. He reported his findings in his amazingly popular book “Beat the Dealer”. If you don’t know or don’t deploy the odds of that card game, you are doomed to suffer endless losses. Ed Thorp is now more commonly recognized as a successful hedge fund manager who delivers rewards annually in the same class as Warren Buffett. Indeed, a few investment wizards do exist and prosper.

John Bogle’s constant reminder that costs matter greatly in the financial casino is also extremely pertinent, especially in an environment whereby the expected near-term rewards are below historical averages. In that casino, Bogle charges that the advisors game is to convert all your money into his money by multiple incremental fees that bolster his payday to the detriment of your end wealth. Many advisors, consultants, media gurus live by their wit and salesmanship, not by their financial wisdom. Fees are wealth depleting lodestones, a lasting money transfer device.

Daniel Kahneman makes the case that these smart professionals know what they don’t know, but perverse incentives are operative. It’s a financial perversion of Donald Rumsfeld’s unknowable unknowns quote. Numerous studies document that financial professionals fail to correctly forecast the future. IPOs are often losers and CEO energized mergers typically lose money for the acquiring firm. Wall Street recognizes these shortcomings by often lowering the stock price of the “successful” acquiring firm just before and immediately after the acquisition happening. The more famous the CEO, the more likely he will stumble with an overreaching acquisition.

Wall Street is dominated by perceived smart money managers who invest gigantic sums mainly for rich institutional agencies. These institutions now constitute 70 % of the daily trading volume and employ the smart money manager cohort. These guys know and deploy statistics in their decision making. Those of us who do not acknowledge the practical need for statistical analysis serve as target punching bags for those who do. Please use statistics when making all your investment decisions.

An overarching conclusion from a host of academic and industry studies illustrate the futility of active investing to produce returns that exceed benchmark Indices, some positive net Alpha. Research and trading costs impose too high a hurdle to overcome. This conclusion is not without exceptions, just that these exceptions are rare.

Superinvestors do exist. Warren Buffett’s 1984 seminar at Columbia University “The Superinvestors of Graham-and-Doddsville" pays homage to the teachings of Benjamin Graham while simultaneously identifying a few members of this select group.

But the odds of having immediate access to this select group are indeed challenging. You need a huge bankroll to invest with George Soros. Given these daunting hurdles, we individual investors should consider passive investing as a viable and acceptable option. Just maybe, market-like returns are not so bad given the rather poor performance records of so many active mutual fund managers.

Just what are those odds? I’ve cobbled together a couple of tables that characterize the odds. The tables should help guide your investment decision making. I call these decision matrices Odds Maps. These maps have many fundamental dimensions.

The first table is titled “The Probability that an Actively Managed Portfolio will Beat Index Funds”. It was lifted from The Tao of Wealth website. That website provided it as a summary from Rick Ferri’s “The Power of Passive Investing” book. In that book, Ferri credits Allan Roth as the primary source of the data set. These things seem to be very well traveled. Here is the data.

1 year 5 years 10 years 25 years
1 active fund 42 % 30 % 23 % 12 %
5 active funds 32 % 18 % 11 % 3 %
10 active funds 25 % 9 % 6 % 1 %

This odds table for actively versus passively managed mutual funds is given as a function of the time dimension and the number of funds dimension. Both influence expectations. In general, they were generated based on countless Monte Carlo simulations completed by Allan Roth. These data are consistent with S&P persistence scorecard studies.

Other studies show similar findings. The specific numbers change a little with each study, but the trendlines are invariant. Eric Kirzer, a Toronto professor, displayed comparable results in a 2000 paper titled “Fact and Fantasy in Index Investing”. Roth published his research in a 2010 Forbes article called “A Winning Fund Doesn’t Equal a Winning Portfolio”.

Rick Ferri did his own analysis that emphasized the asymmetric nature of excess profits from winning actively managed funds and their sub-performing losing counterparts. Overall, portfolio positive Alpha odds increasingly degrade as the number of actively managed components increase and as the timeframe expands.

The keystone input is the baseline rate statistic for a single fund for a single year. Daniel Kahneman strongly underlined the need to always start an assessment with the base rate as a point of departure. Perturb it only if conditions warrant a modification. Remember that bad active managers do disproportionate damage to returns.

Costs are always constant regardless of successes or failures either in the open marketplace or because of portfolio preferences. Burton Malkiel famously remarked that “ Great track records tend not to persist, but high expenses do.”

Depending of your specific preferences (very optimistic or pessimistic) the baseline rate should be adjusted either moderately upward or downward. You should mentally adjust the table to suit you special needs and/or world views.

As a second useful table, I constructed a 2 X 2 matrix that shows the likelihood of positive annual return probabilities in both the managerial style dimension and the trading frequency dimension. The basic data used to complete the table come from both academic work and the DALBAR industry surveys. The probabilities depicted are the likelihood (the odds) of positive annual returns.

Active Management Passive Management
Dynamic Tactical Trading 30 % 60%
Static Strategic Trading 50 % 70 %

Note that I anchored the two extreme input (the dynamic tactical active and the static strategic passive) components using historic data sets. The 70 % odds input is simply the positive annual equity return frequency from market history. The 30 % baseline rate for the dynamic tactical active box (mostly higher trading activity) was lowered below the 42 % value quoted in the first table to honor a more subtle observation that all active managers are not necessarily frequent traders; some active managers hold positions for rather lengthy periods so the baseline rate should mirror that semi-inactive tendency to produce better returns. Frequent trading is hazardous to end wealth.

The other two tabular inputs are my approximations given that the scale ranges from the baseline rates of 30 % to 70 %. Also note that the proposed numbers are simplified to a single digit value. This properly represents the level of accuracy anticipated for these rough inputs. Any attempt to be more precise is illusionary (perhaps even delusional). Again, I recommend any adjustment that echoes your specific circumstances, preferences, or insights.

Additional comments follow in Part 2 of this long posting.

Best Regards.

Comments

Sign In or Register to comment.