The art of calling market tops

Most good books about trading, as well as most famous investors, will tell you: don’t try to call tops or bottoms in the market, it can’t be done. So, they usually advocate scaling in and out of positions to get a good average buy or sell price.

But is that so? Is there really no way, other than watching for clear signs of investors’ euphoria or despair, which tend to accompany major long term highs and lows?

Any “it can’t be done” has always looked like an invitation to try to me. True, you won’t be very successful at calling tops with the standard technical indicators like 200 day MA or RSI or VIX. But of course, big banks and fund managers are not using those standard indicators you can find in any free charting program. They have better stuff. Expecting to beat the market with 50 year old indicators is like hoping to win a Formula 1 race with a 1970s car.
You will also not call tops by listening to the experts on CNBC or other financial news outlets. A news channel survives by showing what its viewers want to see or hear, not by ringing a bell when the market has peaked. They simply ring whatever bells are popular enough on the given day, and that are usually the wrong bells. At the top most viewers want to hear only the good news, because they are fully invested.

The approach I have chosen is to develop my own indicators, using a bit more complicated math than the simple moving averages of yesteryear. I have a bunch of different indicators that have been gradually improved and tested over the years.
Today I want to show how my ELC and BMO indicators can be used for detecting major tops on weekly charts. The ELC is specifically designed to turn up (or down) very close to market peaks (or bottoms). At major long term peaks it tends to show a clear bearish divergence. A picture shows more than a thousand words, so let’s start with the Dow Industrials 1906 peak:

Market made new record highs in early 1906, but the ELC painted a much lower peak. The ensuing 2 year bear market took the index down almost 50%.

By 1916 the market had reached new records again:

A big bearish divergence appeared at the highs in late 1916 and a sharp 1 year bear market followed. In late 1919 the market peaked with another bearish divergence and another 50% decline followed suit.

The roaring 20s showed multiple bearish divergences before the market collapsed:

The highest ELC peak came in 1925 already, a full 4 years before the ultimate peak. The highs in the final advance 1928-29 came with lower and lower ELC peaks. A 90% bear market and great depression followed.

By 1937 the market had recovered a great deal:

But once again the ELC printed a lower high at the peak, leaving a 1 year bearish divergence in place. The market dropped 50% over the next year.

After the second world war a long term bull market started:

Clear bearish divergences appeared in 1955-56. This led to a milder bear market, more a multi-year sideways pattern. Not every bearish divergence in weekly ELC becomes another great depression. But you didn’t miss anything if you stayed out of stocks until 1958.

I will use the broader S&P 500 from the 1960s:

By early 1965 we see a clear bearish divergence. A first dip was overcome, but by early 1966 there was an even bigger divergence. The market crashed 25% within a year.

The bull market resumed, reaching new record highs in late 1968:

But ELC painted a lower peak and the market crashed almost 40% over the next 18 months.

The same scenario in early 1973:

New all time highs for the S&P 500 but a clearly lower ELC reading. A 50% bear market followed suit.

The 1980s burst out into new record territory and by 1987 we saw this:

Once again clear bearish divergences on the chart before the market crashed.

The 1990s massive bull market was very interesting:

Highest ELC value was in early 1996, a full 4 years before the eventual peak just like in the 1920s. We see several bearish divergences that only led to a sideways period or mild pullback. This is something to remember: not every divergence leads to a major bear market, but almost every bear market is preceded by a bearish divergence in weekly ELC.
By early 2000 the market was still making new highs with the ELC making lower and lower peaks and barely keeping above the zero line. The S&P 500 would drop 50% over the next 3 years.

The subsequent advance into 2007 high was a bit unusual:

The highest ELC peak came in early 2004. Several bearish divergences only led to a mild pullback. By late 2007 the market peaked with the ELC down very sharply from its high earlier that year. In fact ELC hardly turned up on that last peak. This showed unusual weakness and a 60% decline followed.

So where are we now?

Bearish divergence in 2014-15 only led to a mild decline into early 2016 lows. The ELC has meanwhile printed new highs for this cycle, so there is no bearish divergence in place at the moment.
This means we have two possibilities. If January 2018 turns out to be a major peak then it will be the first time in over 100 years it peaks out without an ELC divergence. The other option is that this market makes at least one more run to new record highs and then we would probably get that bearish divergence as seen on all those earlier occasions. It will be interesting. Exceptions are there to prove the rule, but if history holds up then it is too early to call for a market top here.
I will keep you posted on how this chart evolves. Here and on my Twitter. And with a bit of luck I will try to call the top when my indicators suggest it is that time again.