What's the most interesting stock-market debate? Value versus growth? Perhaps. But many might vote for fundamental versus technical.
Nobody goes so far as to say fundamental analysis is nonsense. The worst that can be said about it is that the market is often emotional and there's little reason to ignore the trading profits that can be made by riding the coat-tails of admittedly short-term sentiment shifts. Put another way, if one has a chance to make some money buying the shares of an admittedly bad company, why not! We're looking for returns, not brownie points. The same can be said for selling shares of great companies the market may come to foolishly disdain.
The anti-technical crowd doesn't necessarily argue those points. It's objection is more likely to be that they'd love to use it if only it cold really work, but it doesn't. One especially legendary anti-technical challenge is the ripped price-chart exercise. Take a price chart, rip it vertically, show a technician the part containing the earlier trends, and ask him to predict what will happen next. According to the lore, none will accept the challenge knowing the that the answer, contained in the other half of the paper, is right there in the questioner's hand. That's a bit unfair since fundamental expectations cal also return out wrong and since technical systems can be backtested just as fundamental models. But it does, at least, make for a powerful cocktail-party debating point.
Speaking of backtesting, practitioners see that technical models can be shown to succeed, just as is the case with fundamental models. Also consistent with fundamental models, successfully-backtested technical approaches can falter if the underlying conditions change. Those who've tested enough of both kinds of systems know that in reality, the contest between the two approaches is really a level playing field.
That is consistent with common sense. Fundamental expectations drive stock price performance, for better or worse (i.e. as events later show here-and-now consensus expectations to have been reasonable or unreasonable). Technical analysis is a way of systematically analyzing the behavior of those market participants who, presumably, are behaving based on their own fundamental expectations. In other words, the fundamental investors acts based on his own expectations, while the technical investor piggybacks on the consensus expectations of others.
Recognizing this approach as a measure of the sentiment of other market participants, we realize that we need not confine such strategies to the price and volume data that have long been the staple of technical analysis. Any other measure of sentiment fits in.
This screen uses two sentiment gauges, one, a core selection from the technical-analysis repertoire, and the other, data relating to short interest. The specific screening tests used here, applied to a universe consisting of non-OTC stocks priced at or above 5 and with market capitalizations of at least $250 million, are as follows:
- Short covering over the past month ranked in the top half relative to industry peers
- MACD signal indicator ranked in the top half relative to industry peers
- MACD (Moving Average Convergence/Divergence) is the difference between the stock's 26-day exponential moving average and its 12-day moving average
- The signal line is the difference between the current MACD score and the 9-day exponential moving average of MACD scores
- This indicator is a measure of the strength of a stock's trend
Figure 1 and Table 1 show the results of a backtest using all stocks that make the screen, usually numbering in the 500-1,000 range. It covers the period 3/31/01 through the present, and assumes the screen is rerun, and the list refreshed, every four weeks.
This isn't the greatest set of backtest results around. But it is ahead of the market, even after allowing for a particularly rough experience in 2007-08. That's more than many academicians and other passive-investing advocates believe possible, and it was accomplished only with two simple sentiment tests. Even though the screen, by itself, is not the best around, it should at least give pause to those who dismiss sentiment as a legitimate approach to stock selection.
Going further, we'll combine these sentiment tests with the Balanced Fundamentals ranking system. In other words, we'll piggy-back on the positive fundamental expectations formed by others, and at the same time, limit ourselves to stocks that really do seem to merit favorable fundamental assessment. We do this by confining the stock list to the top 10 screen-based stocks, as determined by the Balanced Fundamentals ranking system. A list this size is consistent with what most investors could very comfortably research and/or trade.
Figure 2 and Table 2 show the results of this revised backtest.
Consistent with what one might expect, the combination of sentiment plus fundamentals outperforms the already-market-beating sentiment strategy alone. Comparing down-market results (where sentiment marginally underperformed), we see that adding fundamentals is most valuable during such periods. In any case, the finished model, sentiment plus fundamental, beat the S&P 500 by a wide margin.
For those who would like a greater variety of ideas, the backtest summary in Table 3 shows that we can easily expand the list from 10 stocks to 25.