Hi Guys,
“Things should be made as simple as possible, but not any simpler.” That remarkable aphorism came from the fertile mind of Albert Einstein. The proper balance is critical.
That’s certainly not a modern insight. Leo Tolstoy noted that “There is no greatness where there is not simplicity, goodness, and truth.” Venturing further into history, Leonardo da Vince observed that “Simplicity is the ultimate sophistication.”
Financial modeling has become more complex with the easy access to supercomputer power. The issue is to identify just how much modeling detail is needed to fully capture the essential trends in the forecast. Adding complexity to a model structure by expanding its elements doesn’t always improve the accuracy of the forecast. Many examples of modeling over-specification exist.
One recent problem class illustration is with climate change predictions. Climate researcher James Hansen made rather crude projections as early as 1981; these have subsequently been reasonably validated by worldwide temperature records. The modeling at that time was notoriously incomplete since it omitted major factors (like clouds) that potentially would alter the predictions. By 1988, Hansen improved the climate model fidelity by incorporating three-dimensional atmospheric effects and more physical elements. The 1988 computer model, when tested against out-of-sample recent data, failed to achieve the accuracy achieved by the simpler model.
In many instances, Occam’s razor is a useful concept when confronted with highly complex, interactive problems like equity returns forecasting. One popular form of Occam’s razor quoted in Nate Silver’s “The Signal and the Noise” book is “Other things being equal, a simpler explanation is better than a more complex one.” I firmly support that axiom, and usually take the simplest decision pathway whenever possible.
Following Occam’s guideline, I have consistently searched for simple ways to forecast annual equity returns. Since I am a longer-term investor (in contrast to a speculative agenda), I am quite satisfied with a technique that gets things about right over an extended period (like a multi-year cycle), not necessarily one that has immediate annual accuracy demands attached to it.
For several decades, I’ve practiced the Occam-like approach reported by John Bogle in his wonderful 1999 book titled “Common Sense on Mutual Funds”. Using an Occam’s razor philosophy, Bogle demonstrates the power of that technique over 10-year periods. His formulation relies solely upon only two components: a fundamental element and a speculative element. Its correlation with equity market data is remarkable.
The fundamental element has two sub-contributors: a corporate profit growth estimate and a dividend yield estimate. The speculative element is interpreted as a measure of market participant emotional zeal. It usually is characterized by the market’s current P/E ratio relative to its historic average. The speculative assessment adopts a return-to-the-mean mentality.
To forecast upcoming equity market rewards, the various components in Bogle’s approach must be estimated for next year, and then updated annually. Here is how I do that forecasting.
For the fundamental component, I evaluate the corporate profit growth rate by estimating the US real (after inflation) GDP growth rate. Analytical work that I completed years ago show a tight correlation between corporate profits and GDP growth. The correlation is that Corporate Profits equal 1.7 times the estimated GDP Growth rate.
GDP growth rate has been historically composed of mostly two basic factors: population growth and productivity growth. Again, historically in the US, population growth is about 1 % annually and productivity growth has been about 2 % per year. You get to pick your own values by assessing current macroeconomic factors. If you feel Apple will introduce a new technology and/or the auto industry will invent a new rechargeable battery system, you should upgrade your productivity estimate. It’s always your call.
Historically, dividend yield has been much higher than today’s 2 % rate. Again, it’s your call. I often input an estimate close to the current level unless I anticipate a jump in inflation rate.
Now for the challenging task of estimating market participant emotional behavior. I have never had a comfortable handle on this illusive parameter. The Behavioral research community has contributed some very generic guidance, but few actionable rules in this arena.
But Nate Silver’s book provides a deployable clue. In his chapter that discusses equity markets (“If You Can’t Beat’em”), Silver shows a series of data sets that highlight equity returns to P/E ratios. By curve fitting the longer-term data, an algebraic equation can be developed that shows a linear influence relationship.
If you assume that the regression-to-the-mean hypothesis holds, and allow for a 50 % correction overshoot (a nod to Behavioral studies of overreaction), a best fit equation can be derived to evaluate the P/E adjustment needed to estimate the speculative contribution potential on an annual basis.
If a baseline P/E ratio of 15 is postulated (reasonable, perhaps a little low), then the speculative component is positive for an undervalued market, and the incremental speculative contribution is 0.53 times the factor (22.5 - current P/E). For an overvalued market, the incremental speculative component is negative and is 0.53 times the factor (11.0 - current P/E).
Real expected equity return is simply the sum of the three components identified.
Actual Return is Real Return plus an inflation estimate. Today, inflation rate is about 1.7 % annually. Historically, it has been at least double that figure. Again, it’s your task to make the call. I anticipate a moderate increase.
Now for the mandatory disclaimer and some obvious reservations about my analysis.
I have not checked this exploratory formulation against any actual data. That’s an arduous task needing completion. I may never do that task because I am a committed equity investor who plans to always hold some equity positions in a diversified portfolio. My incentives are minimal given by cautious and incremental approach to portfolio modification. I’m sure your incentives differ sharply and logically from mine.
The method does enjoy the merit that it is in substantial agreement with that proposed and tested by John Bogle. I surely do not expect this analysis to be absolutely correct for any given year. It is long-term macro by design. But it does have the prospect of being approximately correct over more extended timeframes, like a 10-year cycle. This approach is definitely not for day-traders or short-term speculations.
As always, buyers beware of any "How Wall Street Works" models.
I do not recommend that you act on my formulation. I present it to this forum to simply inform. It just might help you make some portfolio adjustment decisions. It might not. It does suggest that expectations for outsized long-term returns are highly unlikely.
The markets are dynamic entities in perpetual motion. The marketplace participation has shifted from mostly individual to institutional dominant. Trading volume has increased exponentially. Position holding periods have decreased by an order of magnitude. Media hype is everywhere. All this makes models and modeling more vulnerable. Always remain skeptical.
Best Regards.
Comments
"The fundamental element has two sub-contributors: a corporate profit growth estimate and a dividend yield estimate."
"The speculative element is interpreted as a measure of market participant emotional zeal."
Very much appreciate.
"Always remain skeptical."
Indeed, our best engineers (and financial analysts) always remain skeptical.
The irony is, our best program managers (and traders) can not. They've got to believe.
Thanks for sharing MJG.
Keep it simple is within the bounds of knowledge and experience of the individual.
You noted: What you have noted here may not be valid for many to understand or perceive. Or perhaps another methodology is more simple for another. I am sure you would agree there may be many pathways to a given goal related to investments.
I often wonder how many of the money experts, advisors or individual investors from 5 years ago, this month of October; were using any of the most simple forms of monitoring the broad markets, particular sectors or funds where they had monies invested. Basic 50/200 day moving averages and another basic measure of relative strength between a low of 30 and a high of 70 could have been guides for caution and to alert one to maintain a close watch of monies.
As to estimating growth or earnings of the broad economy or company; I will have to leave that to others. I will again suggest that our investing world that is perverted by central bank policies, political fighting and public and private debt burdens; remains upon unstable ground in many areas.
What passes today for a fundamental aspect of the markets is difficult to guage. My recall notes that some accounting standards for business has been adjusted during the past few years. If this is the case; any company beating earnings estimates by a penny may be found via accounting methods; which may or may not be adjusted in the next reporting period. But, there are those who may cheer such a positive earnings event to wave their own flag of victory for their being in the right place at the right time; at least for the moment.
As for the behavioral aspect, of at least the retail investor; I suspect few of us here are surprised by current feelings of those who remain a bit skeptical of the investment marketplace. The overall perception has at least some basis in the separation of "classes" of retail investors, from the trading desks. A recent report (can't recall the source) indicated that 70% of the trading activity is from the machines and the algo's. I believe many investors, who are not formally trained/educated students of the markets; find a massive shift in "how" to invest that not longer lives in the fundamentals world; but in the machine world.
I will note that the "luck factor" of being in the right place at the right time may have more weigh today; versus 5-10 years ago.
July 25, 2012 indicates a possible shift or at least a marker day/week for our house. The U.S. dollar index topped after a sustained upward move, safe haven bonds of the U.S., Germany, the U.K. and Japan had a reversal of pricing/yields, and the U.S. 10 year note; which is a benchmark for many rate watchers is attempting to rise above 1.8%. Whether any of this is meaningful at this date is not yet established; but has caused our house to watch more closely.
The simple side is how to best use whatever one may find of understanding and value to measure their investments and how the markets may impact their investment mix.
Our house is surely down on the list of skills with complex number sets and systems to determine buy and sell points for our holdings; much of which is affected by a limited amount of time. We always attempt to do our best, regardless.
My words and thoughts have wandered off path, with this reply; but perhaps there is something valid, for someone's consideration. This is a very complex subject area and I surely have not done justice, with the topic.
Take care,
Catch
Thanks for your thoughtful and thought-provoking post. I enjoyed it, although some of the logic that guided your detailed interpretations did escape me.
Indeed in the last decade, institutional investors have replaced individual investors as the primary market movers. And their trading frequency and volume have furiously increased. Supercomputers have greatly facilitated this escalation. Huge trades are now completed in microseconds.
I’m more ambivalent over this emerging development than many others are. Both positive and negative aspects to this rapid trading exist.
The super machines permit rapid fire analysis that exhaustively examines a nearly endless array of parameters in the decision making assessment. Evaluation errors are not made. The machine is not prone to the common human failures coupled to hubris and overconfidence. It is not lazy and will not need frequent coffee breaks.
The downside is that the machine is captive to the limitations of its programming. Hopefully all programming errors have been discovered and rooted out of the system. The machine has no imagination, and will slavishly follow its instruction set, even when it is no longer appropriate given changing market conditions and prospects. Human intervention will always be needed to monitor for Black Swan events.
Overall I do not fear computers or the ascendancy of institutional investors to their primacy role. At present, there is a sufficient independent institutional population to secure the diversity of opinion needed for a functioning mostly Efficient Market. Independent judgments are mandatory to avoid the pitfalls of group-think.
I too like and use (along with a small set of other indicators) Index moving averages to guide equity entry and exit points. Since I am a long-haul investor my signal set tends to have whiskers in terms of timeframe.
The list of candidate indicators is never-ending; each has its merits and its shortcomings. It is likely that an indicator soup composed of several fundamental indicators serves the investor best. For example, one popular speculation-influenced indicator is the AAII Sentiment survey reported on their website. It is interpreted as a contrarians signal To each his own technical poison.
I was fully puzzled by your selection of the late July 2012 date as a significant signal for the US Dollar Index (DXY). My perplexity is not surprising. To me, it looked like daily noise. I do not comprehend its complex construction (it is not clear why the dominant European component and the lesser Japanese component should be in sync) or its historical performance (lots of peaks and valleys lacking fundamental economic drivers). The DXY Index baffles me.
It is true that the DXY registered a local mini-maximum in the money exchange market landscape, but so what? Many such local maximums were recorded in the last three years. One such local high trading mark was logged around June 1st of this year. Catch, why is the July peak so relevant to your thinking and planning? Your post suggested it was a particularly vital tipping point.
Thanks again for your stimulating contribution. It certainly expands the discussion framework. It motivates participation, and that should improve everyone’s investment decision making.
Best Wishes.