How Helpful Were Market-Maker Forecasts in 2013?
In the year 2013 we were able to discern from their hedging actions just what price ranges the market-making [MM] community thought were likely for 2,393 stocks and ETFs at the beginning of the year, and for 2,475 of them by the end. Over the course of the year, some 600,000+ forecasts were implied, day by day.
Here's the short story: Out of those forecasts some 240 were selected, ex-ante, at the time of the forecast, as choice investment candidates to pursue a wealth-building objective, and were published in articles on Seeking Alpha. By a previously-defined, standardized published score-keeping strategy, their subsequent price change experiences were recorded and averaged, as follows: Of the 240 selections, by year-end 205 had reached sell targets or position holding time limits and had been closed out, with 35 remaining to meet decisive action.
Of those 205 positions closed out, 84% were profitable, and including the losers, a geometric average gain of +6.5% of the capital commitments was earned, requiring an average holding period for all 205 of 50 calendar days, or a day over 7 weeks. When those gains are reduced to daily gain rates per position and then expanded to an annual average for the 205, the annual rate of gain is at +58%. The average of maximum price drawdowns from cost for the 205 during the periods held was -5 ¼%.
This is not a measurement of the gain of any specific investment portfolio and does not follow CFA GIPS standards for portfolio measurement because that is not its purpose. The point here is that the annual rate calculated is intended to describe a rate of gain widely available to a large number of investment vehicles at various points in time throughout the year, held for varying lengths of time under a predefined opportunity strategy.
It presents a pattern that under a similar strategy may be highly productive in a year where market surroundings are similar to the year 2013. The notion of advantage over market behavior (a.k.a. alpha) is believed to be proportionate under less favorable conditions but not proven (yet) in times of stress.
Details of the measurement
We wrote some 85 articles on Seeking Alpha describing their expectations. In many of the articles detailed tables of several stocks or ETFs were presented, showing price limits within which MMs felt it necessary to protect themselves from adverse price moves, both higher and lower. Over 700 specific price range forecasts were listed.
Our articles noted that if one took the upper end of those implied price ranges as sell targets for long positions in the specific stocks, a score could be kept, based upon the actual subsequent market price movements. A pre-defined standardized strategy to limit holding patience times 3 months (63 market days) made capital commitments time-efficient. Position entry prices at close of market on the day following the forecast and position exit prices at end of day first meeting or exceeding the sell target, or at the 63rd day, whichever came first, defined the % payoffs and the number of days the position was held.
The articles noted which stocks or ETFs at the time of forecast, ex-ante, were held to be most attractive in achieving a wealth-building objective. Those distinctions were derived by applying the scorekeeping strategy described in the paragraph above to prior forecast instances similar to those in the then current forecast. The forecast standard used for each subject was the Range Index [RI], a measure that tells what percentage of the forecast price range lies below the market price at end of day [eod] of the forecast. The RI in effect splits the forecast into upside and downside prospective price change potentials.
Most articles, with few exceptions, noted for each subject the number of days in which a forecast had previously been derived at any level of Range Index (the subject forecast population size), and a count of those at a similar level to the current RI (the sample size). Subjects were screened to exclude those with forecast populations of less than 3 years (756 market days of 252 per year) and those with fewer than 20 sample forecasts.
To develop a basis for preferencing investment selection candidates within each article, the sample prior forecasts for each subject covered in the article were evaluated as winners or losers and their Win Odds were recorded. By article, those subjects passing the previously mentioned screens were further screened to eliminate any with Win Odds less than 7 out of 8 or 87 out of 100. The same analysis produced for each subject an average net gain among all of that subject's sample (% payoff), and a measure of the average market days held by the sample positions. From that, for ease of subject comparisons within the article, annual rates of net gain were computed.
From each article containing price range forecasts, some number of specific investment candidate subjects were selected as either groups of "best", "selected", "top", or other designation, or were referred to by SA article editors as either "about" or "included." From the over 700 instances of forecasts in the 85 articles, 240 specific stock or ETF subjects were so named.
The same strategy used to provide a basis for selection of the 240 subjects was then, ex-post, applied to each of them to determine the results previously described at the beginning of this review article.
Restated, by year end 205 of the 240 had reached closeouts; 84% of those were profitable, and including loss positions, the average gain was +6.5%, took 50 calendar days of capital commitments on average, and provided average annual rates of gain of +58%. At year-end the 35 positions still open had only 5 at a price below its entry price, and the group of 35 showed a geometric mean net gain of +5%. The shortened holding periods fore-closed by year end would produce in several cases a high annual rate of gain inappropriate for comparison with those that were taken to the strategy's completions.
The proportion of still-open positions in profitable condition, and the size of early gain to year-end suggest that these positions would not seriously bias the results of the completed positions.
What could a portfolio do with specific recommendations?
Big annual rates of gain from a couple of hundred examples are hard for many folks to imagine, built from +5% or +6% one at a time, six or seven times in a year. How can that compound into real wealth building when losses are included?
To answer that question we decided to create a few specific examples using the recommendations made in the detailed .pdf file. Since there were seven featured forecasts on the last market day of 2012, December 28th, we started there and built seven illustrative threads for a portfolio.
Each thread started with one of the seven choices and proceeded along the standard strategy, reinvesting the proceeds of the closed-out prior position in another recommendation as soon as another selection that had not been taken by a parallel thread was available. Each thread was engineered to have at least one losing position that consumed one quarter of the year. All of each thread's other selections came from the remaining 9 months of the year, without any overlap in calendar sequence or duplication between threads. Dividend payments were ignored.
No one thread was stacked to grab all the best opportunities, but many of the very productive experiences did get included in one or another. We didn't want you to get bored looking at dozens of ho-hum results in size, just because they only took a week or two to hatch.
Here is a summary of the seven threads, with illustrative hypothetical investment commitments of $10,000 each on January 2, 2013:
Thread #1 Starting with TQQQ, after 11 positions with a loss in AAPL, ending at $34,893
Thread #2 Starting with UPRO, after 7 positions with a different loss in AAPL, ending at $30,441
Thread #3 Starting with UYM, after 9 positions with a loss in AMZN, ending at $24,530
Thread #4 Starting with ROM, after 9 positions with a big loss in AAPL, ending at $17,683
Thread #5 starting with NDX, after 7 positions with a loss and gain in SCO, ending at $17,814
Thread #6 starting with a loss in SGOL and after 4 other positions ending at $17,588
Thread #7 starting with a big loss in AGQ and after 6 other positions ending at $18,097
The initial investment capital of $70,000 could grow to $142,948, a little better than a double, after current-day transaction costs on 55 total round trip transactions in the course of a year. Of course, if the average of all 205 selection experiences could have fitted perfectly together at their average +58% annual rate of return for the year, the $70,000 would have grown to only $110,600. If invested in SPY the $70,000 would at year end have had a value of $88,563.12
Why all this bragging?
An often-quoted person once stated "If ya can do it, it ain't braggin'!"
But more importantly, our position is that Market-Maker insights into the every-day activity of equities markets provide a valuable guide to wealth-building strategies. We have seen this borne out in our own experience year after year. Now this record offers numerous examples that have been published before the fact, across a meaningful period of time, and have subsequent real price actions verifying that notion.
The supporting data is in a .pdf file accessible for public view and audit at http://www.blockdesk.com/2013report/2013perf.jpg.
It all goes to support our contention that conventional investment strategies of buy and hold force on investors returns that are sub-optimal and endanger them with often irretrievable risks of loss of investment time.
Well-designed, simple investment strategies can force reexamination of the prospects for all relevant investment candidates. This is done by requiring each investment to have a specific, unchangeable sell target that must be honored when first achieved, and all positions to be otherwise limited to a common maximum patience length. As reinvestment funds are made available by mandatory liquidation of position holdings at sell target achievement or holding period limits, each occasion presents the need for a fresh reappraisal of available qualified investment candidates.
We believe that such strategies permit uninhibited, frequent reexamination of current-day forecasts of what can realistically be expected, necessary to the achievement of return-vs.-risk tradeoffs most satisfying to informed, demanding investors.
Further, we believe that investments should be diversified not only among capital commitments, but among time-period investments. A discipline which creates this naturally is one that requires having specific, unchanging price targets and holding time limits for each capital commitment separately. The reinvestment of each subject's capital and time commitment, triggered by the eventuality of that discipline, forces a diverse-in-time series of actions, similar to expiration-ladder investing in bonds.
Today's state of markets competition and information technology have reduced transaction costs to a trivial level where reinvestment candidate competitions deserve to work from a clean slate of cash positions as a starting point. This insures that the recent success (or failure) producing the reinvestment need allows the prior subject equal footing with all the other candidates.
Such a discipline, accompanied by a score-keeping practice of how well sell targets are met, and how prices behave as holding periods approach and pass by, provide an opportunity to refine the specifics driving the discipline, making for possible strategic or tactical improvements.
If the "alpha" opportunities were as small as we often see offered in many academic or professional journal articles, we would likely not be moved to put forth such an effort as is reflected by this article. But the true differences are enormous, and the investing public deserves to know what can be done - often by investors themselves - to significantly improve their investment results.
We hope we can contribute to that end, and intend to continue to offer articles on Seeking Alpha from time to time providing market-maker-derived inputs.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.