When Seasonality Can Be Misleading
Never take a trade just based on seasonal patterns, warns Larry Williams, who instead uses them to confirm or strengthen signals given by charts and volume.
My guest today is Larry Williams, and Larry, you’ve been doing seasonal analysis for longer than most of us have been trading.
I wrote the first book ever on seasonality in stock and commodity prices way back in 1973, and now it’s everywhere; everyone’s into seasonals. I might not be the best with it, but I’ve used it the longest, so I have a little insight into it.
I’m anxious to gain that insight. Do you see some applications of seasonal analysis that you think are wrong?
Well, I think the big thing is that it is not mechanical. A seasonal pattern is suggestive. It’s not mandatory that it’s going to happen; it suggests we might get a rally at that point.
And, they change. Right now, the big topic is “Sell in May and Walk Away.” Well, in actuality, the best seasonal pattern from 1900-1950 was to buy in May. Now it’s changed from 1950 on and it was best to sell in May and walk away.
You really have to monitor these things as they come and flow in and flow out. It’s not just a mechanical trade of a seasonal pattern. You can’t do that and make money.
I’ve written about it myself and I usually refer to it as a “tendency.” It’s nice to look at it when it dovetails with my other technical work and volume studies; otherwise, it’s irrelevant.
Absolutely, it confirms. Let’s say I have two trades, and one has a seasonal and one doesn’t: I’m going to go with the one that has a seasonal.
So, it confirms a trade, but I won’t take a trade just because of a seasonal pattern. I used to do that and it was not successful, so I learned. When I get my fingers burned, I go, “Oh, don’t do that!”
Yeah, that can be really hazardous because you’re going against all of the other evidence.
Yeah, and I learned the hard way.
Well, we have that in common, Larry. You’ve seen a lot of technology changes and changes in the industry. Are there any that you think make it more difficult for traders?
I think the big thing traders are missing out on right now, and it’s on a very short-term basis, is that most are still stuck in the time frame charts.
We’re trading the markets basically 24 hours a day now. So you have somebody using a time period when the market was inactive and trying to get a trend when the market is active with the same time frame.
So the time frame trading, I think, is dead, because you have these hours where nothing takes place, but that gives the same impact, the same trend data as when the market is active.
I think a volume chart or a tick chart is far superior to the old time charts.
Yeah, a lot of people do use tick charts. I’ve never played around with them that much, but I find it interesting that some futures traders will use a certain number of ticks for some markets and a different number of ticks in other markets. Do you think that’s valid?
I think it’s goofy. I think what really matters is how many swings you want to catch during the day. If you’re a one-minute trader and you want to catch 150 swings during the day, you can find how many ticks create that action in the marketplace.
I’m an old man; I’m 70 years old. I want a market that doesn’t have a jillion trades up and down during the day. So in the S&P, I like about a 10,000-tick chart, and that gives me maybe three or four swings during the day. If you want more than that, you go down to a slower number.
There’s no magic number to any of this. All of these people claim magic; there is no magic number to anything in the market. From my experience in 50 years of trading, you want to see what fits your trading personality.