The Secret of Cycles

You already know I love cycles… but I certainly don’t swear by them.

Sure I keep a close watch on them, and I use various tools – including one particular algorithm (that I keep tightly under wraps) – to tap into the cyclical rhythms of the market.

But I’ll admit, cycles aren’t as clean as some would suggest. In fact, there are many challenges traders face when they first wade into the world of cycles.

For one, there are too many of them. You’ve got the Presidential and Decennial cycles that Harry wrote about. But you also have the business cycle… a 30-year commodity cycle… the lunar cycle… the 45- to 60-year Kondratieff Wave.

Wait! Did I just give you a 15 year range on the frequency of the Kondratieff cycle?! Yes – and therein lies the challenge. Which cycles should you watch?

There is also the issue of variability in the time span between repeating peaks and valleys. In other words, you can’t set your watch by cycles.

It’s human nature to yearnrn for predictability. That’s why the idea of a perfect cycle – for example, one that peaks with precision every 10 days – is so appealing. And many traders search high and low for the Holy Grail of cycles that will guarantee 100% accurate market timing.

Alas, that Holy Grail doesn’t exist.

Here’s a chart of the S&P 500. At the bottom of the chart is a fixed-period cycle indicator. The period describes how many days pass from one cycle low to the next. This cycle is set to repeat every 10 days.

See larger image

You’ll see some of the 10-day cycle lows actually match up perfectly with price lows in the chart. Others don’t, which is why it’s foolish to arbitrarily follow only one fixed-period cycle to make buy/sell decisions.

I’ve found that cycles naturally expand and contract. That means you may observe a period of time in which mini-cycle peaks occur every 10 days or so. Over time, the period between cycle peaks is stretched out to 15 or so days (showing expansion). But then, as more time passes, the dominant cycle may contract so that only five days separate subsequent peaks (showing contraction).

One way to mitigate the unreliable variability in tracking just one cycle is to track three or four cycles. Sometimes the various cycles will give you conflicting information. That’s fine, you simply factor the lack of agreement into your analysis… looking for further evidence elsewhere.

But other times, multiple cycles will all agree, or show confluence. This is valuable information. If you see three cycles all forecasting a market low within a day or two of each other, that’s a strong signal you’d better not ignore.

As with the market in general, there’s both an art and a science to cycles. Only fools set their watches by market cycles, just as only fools ignore them altogether.