Option Investor
Educational Article

Probability, Empirical Evidence, and Wives Tales

Printer friendly version

Ever heard of the Hirsch Organization? They are the folks that publish the Stock Traders Almanac, a great publication chock full of historical information and market statistics. From a recent Forbes article, I learned that in the current issue of the Almanac Investor newsletter, Jeffrey Hirsch noted that September has historically been bad for stocks. "When portfolio managers get back from the Hamptons after Labor Day, they tend to clean house. It has been the worst month of the year since 1971 for the Dow, S&P and NASDAQ, averaging monthly losses of 1.3%, 1.0% and 0.8%, respectively," writes Hirsch. "Basically, avoiding September can be beneficial to your portfolio."

But that's a coin toss at best. According to the Forbes piece, The Almanac also suggests that the September-to-January period is a good time to be invested in pharmaceutical, consumer and telecom stocks, and recommends buying exchange-traded funds in each of these sectors. Again, repeating the Forbes article, the Almanac Investor recommends Pharmaceutical HOLDRS (amex: PPH) and Dow Jones Healthcare iShares (amex: IYH) in pharmaceuticals; in consumer issues, it flogs Consumer SPDRs (amex: XLY) or iShares Dow Jones Consumer Cyclical (amex: IYC) or Non-Cyclical (amex: IYK); in telecom, it suggests to go for Telecom HOLDRS (amex: TTH) or Telecom iShares (amex: IYZ). How about semiconductor stocks? Avoid them, as September has been a historically bad month for them.

Seems market history is all about seasonal timing. The Forbes article also goes on to point out, "Sy Harding of Street Smart Report, uses what he calls his 'Seasonal Timing Strategy' which says to exit the market on the 4th trading day of May and to buy back in on the second to last trading day of October. Harding uses MACD to confirm his entry and exit points so that he doesn't sell a rising trend or buy into a declining one. As far as this particular September goes, Harding is mixed. He believes the 'bear market rally' could extend another two weeks, but that by the fall, markets will retest the July 23 low."

Want more? Quoted again in Forbes, "James Dines, editor of the Dines Letter, puts at least some stock in seasonality. Dines turned from bull to bear last week based on what he sees as too much bullishness as well as data showing that, since 1961, down Septembers outnumber up Septembers by 2-to-1. Dines also makes an interesting observation about Labor Day: 'If the market declines in the four-day week following Labor Day, one should postpone buying for one month. It worked splendidly in 1994, 1998, 1999, 2000 and 2001, when postponed buying provided buyers with prices near the bottom in all five Octobers. On the other hand, if there is a gain in that four-day week, buy because the market will probably keep going higher.'"

Interesting observation to which I'll pay attention beginning next Tuesday.

Conversely, Forbes wraps it up with the following contrary opinion: "Marvin Appel at Systems and Forecasts dismisses the importance of September's bad rap. 'To the extent that seasonality matters, you make 1% less in September,' says Appel. 'When markets swing 2% to 3% a day, this doesn't seem significant.' Appel is bullish and sees the broader markets gaining another 10% through the end of the year."

Confused yet? Don't be. On a statistical basis (Come to think of it, Mark Twain had a great opinion on statistics. "There are three kinds of lies: lies, damned lies, and statistics.") there is the probability to be correct a greater percentage of the time.

However, getting caught in the statistical minority will land us in the poor house with the wrong bet or too large a bet. Again, the only lesson I take from this is to acknowledge the statistics but don't wrap your trade in statistical dogma. Sometimes, we are going to wrong. It's the nature of the game. Better to cut our losses if we are wrong and wait for a better trade.

In that regard, trading is just like baseball. We wait for the right pitch before taking a swing. Whiffing a trade is a lot more financially painful than whiffing a baseball for a strike. But as Warren Buffet points out, there are no "balls" and no penalties in refraining from swinging at a lousy pitch.

Where am I going with this? To dinner, of course. Please pass the crow and help me get my foot out of my mouth. Seems that last weeks expectation that the market would remain temporarily bullish was flawed. It's pretty difficult to demonstrate bullishness when the S&P 500 fell from 962 to 917 in one week. While I pointed out that the stochastic was due to fall (it did) and the candles ought to theoretically fall to support lines (they did, and then some), my stochastic timing and candle lengths were off. I was anticipating that the next bullish trade would come when stochastics bottomed in harmony with candles hitting support lines. I missed that by a long shot. While there was money to be made on Monday when the SPX bounced at 930 (direct hit!) all the way to 950, 950 proved unsustainable, from which the index fell for the rest of the week.

Can I still hold the notion that we're seeing a temporary bull within a primary bear, a.k.a. bear market rally? Maybe. But the greater weakness than expected certainly caught me off guard. The markets were failing to catch a bid. Perhaps we could dismiss this as just the market falling under its own weight from lack of buyers, as many seem anxious to start the holiday weekend early. Lack of volume would seem to confirm that. However, I was uneasy about the pickup in selling volume yesterday compared to buying volume. Overall, the rally seen since July's lows has been "eyebrow raisingly" low. That is not the stuff of sustainable bullish movement, and in no way represents the beginning of a secular bull market.

So what now? Do we stick with the cyclical bull theory within the secular bear market? Or did July's rise not even count telling us that the market merely encountered a short-lived relief rally? I don't have the answer.

But Consider this. Richard Russell, 78-year-old grizzled market veteran of 60 +/- years, and publisher of the Dow Theory Letters since 1958, notes that corrections happen in much faster timeframes than the primary trend. Thus, if it took from March through July -four months - for the Dow to sink from its high to new lows, but shot upward for a 50% recovery in just one month, we're still in a secular bear market. The rate of market recovery from the lows came twice as fast as the decline. Ergo. . .bear market. That said, we must not trust bullish moves to last long, or at least until they last longer than market declines.

On the other hand, what if we cheat a little bit and engage in some rearview mirror analysis of the tealeaves, err, I mean charts? Can a case still be made for a bit more endurance of the erstwhile cyclical bull? I mean, if volumes are low, shouldn't the return of vacationers next week pop the markets out of this week's funk? That will happen, right? Charts please!

S&P 500 daily chart

Here's to the power of "fitting" trendlines to market action that has already occurred. I sometimes wish I'd never learned that from Jeff Bailey, as it has given me, and I'm sure many other traders, and excuse to stick with a trade when our expected trendlines or retracements, or averages didn't pan out as we originally thought. The fact is, it's a great tool that opens our eyes to other factors that our biased eyes would not see in the heat of battle. It helps for "next time".

So what can we learn here? Maybe our red trendlines we drew last week were off. Well, yesterday, SPX bounced (barely) at the horizontal trend line, a line for former resistance. Since it held, I considered giving myself another chance to be correct with a bullish trend. After all, I wasn't completely wrong. . .yet. But I did note the close under the 50-dma (magenta line), which is no pillar of strength. Then again, the 5-period stochastic was nearly oversold. "Will it bounce?" goes through my head.

Better yet, how about I just redraw my trendline to "fit" the current action. Yeah, that's it. Forget the wick at July's low and just connect the body of the candles (solid blue line)! How about that! A perfect fit at today's low almost exactly at 10 a.m. ET! (Remember, today's candle that is now a doji was long and red at that moment in time.) Maybe that will hold as support and the new, blue trendline is really more valuable under the current circumstances than the old, red one. Anyway, at this moment, it really seems to fit better. That new trendline may actually be pivotal buy program entry for a number of large traders. It would certainly explain why the whole market caught a bid at that time.

I might also now note that the candle now rests right at its 50 dma as of today's close, which may provide support, as the stochastic is nearing oversold. Perhaps I'll be safe remaining cyclically bullish at this level.

Still, the other side of the stochastic coin is that it has not hit oversold yet, is still in fairly steep decline, and buyers are gone on an early vacation. I'm not likely going to see my bullish trade make "huge coinage" tomorrow.

Alright, so I justified staying bullish based on the blue trend line. But I still have an itchy trigger finger. That was my line of last resort and I have no other possible bullish lines to "fit". While I can create lines to fit candles, I have to stop short of creating candles to justify my lines! If that ever happens, we'll begin publishing the until now secret "Hallucination Trader (.com)"! The dominant financial news media already has plenty of analyst material available to make it a rousing success.

All kidding aside, to be blunt, I have no idea if the bulls can grab the ball back from here and move the candles up from here for another stochastic run up the daily chart. It is certainly possible based on the new trendline and support at the 50-dma. On the other hand, that support line could fall under its own weight, which would be decidedly bearish and would keep the stochastic pointed down. That will be a technical violation that I think would hinder recovery. Still, I can't put much faith in the move in either direction as long as absent players keep the volume low.

From all this, the only conclusion I can draw is that I won't see the pitch I'm looking for tomorrow. So I'll likely sit out of tomorrow's action - remember, no penalties for failing to swing at bad pitches. A friendly reminder that probability, empirical evidence, and wives tales are never an exact science.

Make a great weekend for yourselves. See you next week!

Options 101 Archives