Major indexes have failed to rebound from lows, suggesting weakness The U.S. stock market’s major trend now is down, so act accordingly. That’s what the Dow Theory, the oldest stock market timing system that remains in widespread use today, is saying. It was created a century ago by William Peter Hamilton, who at the time was the editor of the Wall Street Journal. He introduced his theory in dribs and drabs over the first decades of the 20th century on that paper’s editorial page. Though not all adherents of the Dow Theory see eye to eye on how to translate Hamilton’s editorials into specific market-timing rules, they generally agree that the market must jump over three hurdles before a “sell” signal is generated. They are: Hurdle 1: Both the Dow Jones Industrial Average and the Dow Jones Transportation Average must undergo a “significant” correction from new highs.Hurdle 2: In their subsequent “significant” rally attempt following that correction, either one or both must fail to rise above their pre-correction highs.Hurdle 3: Both averages must then drop below their respective correction lows. In the opinion of two of the three Dow theorists monitored by the Hulbert Financial Digest, the first of these three hurdles was satisfied by the market’s decline from its late-December highs to mid-January lows — which took 4.1% off the Dow Industrials and 6.1% off the Dow Transports. Those two Dow theorists are Jack Schannep, editor of TheDowTheory.Com, and Richard Moroney of Dow Theory Forecasts. That set the Dow Theory clock running on the remaining of those hurdles. The second was met in the market’s rally that began from those mid-January lows, since neither index was able to surpass its late-December highs. And the final hurdle was cleared last week, with the Dow Industrials DJIA, +1.14% on Jan. 28 closing below their mid-January lows, and the Transports DJT, +1.44% doing so two days later at Friday’s close. By the way, it will take more than just a few days of strong market action for the Dow Theory to reverse this “sell” signal. That’s worth remembering in the wake of the stock market’s impressive rebound on Monday of this week, when the Dow Industrials soared 196 points. In fact, according to a strict reading of the Dow Theory, the market must jump over three more hurdles to generate a “buy” signal — hurdles that are just the reverse of the ones it just cleared when generating the “sell” signal. I say “strict reading” of the Dow Theory because Schannep follows a variation of it that focuses on the S&P 500 Index in addition to the Dow Industrials and Dow Transports. And because the S&P 500 last week did not break below its mid-January low of 1,992.67, and closed Monday 1.4% above that level, Schannep remains on a “buy” signal. However, if the S&P 500 in coming days does end up below its mid-January closing low, then even Schannep’s variant of the Dow Theory would generate a “sell” signal. What about the third Dow theorist the Hulbert Financial Digest monitors? He is Richard Russell, editor of Dow Theory Letters. You may recall that, on many other occasions in recent years, Russell has been very bearish on stocks while Schannep and Moroney have been bullish. The situation now appears to be reversed, as Russell is now forecasting an “upward explosion in U.S. equities.” However, it’s not clear if the Dow Theory plays any role in his optimism. For example, he recently wrote to clients that his forecast is based on the “new trillions in fiat money” that central banks are creating. This huge disagreement between Russell and two fellow Dow theorists might incline you to dismiss the approach altogether. But that might be unjustified. Consider a study conducted in the mid-1990s by three finance professors — Stephen J. Brown of New York University, William Goetzmann of Yale University and Alok Kumar of the University of Miami. They fed Hamilton’s market-timing editorials from the early decades of the past century into neural networks, a type of artificial-intelligence software that can be “trained” to detect patterns. Upon testing this neural network version of the Dow Theory over the nearly 70-year period from 1930 to the end of 1997, they found that it beat a buy-and-hold strategy by an annual average of 4.4 percentage points. Their study appeared in the August 1998 Journal of Finance. Mark Hulbert Dow Theory sell signal?Both Dow Averages recently failed to surpass their late-2014 highs, and then last week brokebelow their mid-January lows.