Monday, September 14, 2015

Dow Theory Special Issue: Lessons learned from the Chinese stock market crash (II)





How to exit when the Dow Theory refuses to setup “orderly” primary bear market signals.




Recently, I wrote about the primary bear market signal for Chinese stocks.


As you well know this primary bear market signal came much too late, if we are to judge from the percentage decline from the top to the exit point.

Some days before the primary bear market signal for Chinese stocks, I raised a thorny issue: What to do on the rare occasions (not yet seen in US stocks in +115 years) when the setup for a primary bear market signal (that is secondary bearish reaction followed by a rally and subsequent violation of the secondary reaction lows  “type 1 signal” or outright violation of the lows of the last competed secondary reaction “type 2 signal”) fails to materialize at safe levels? By “safe," I mean an exit point (lows of the secondary reaction to be violated) close enough to the last recorded top, which helps us exit in an orderly manner. More specifically: A good exit level is, i.e. a loss/decline of ca. 5-10% from the top. A bad exit level is a loss/decline of -30% from the top as we have recently seen with Chinese stocks. You are encouraged to read my first post on this issue. 

Below is a rough depiction of what I am talking about:


This is what the Chinese stock market recent did: A very rare occurence from which we can learn


I said that Chinese markets, unlike US markets (crashes of 1929, 1987 and 2008, included, which allowed for "good" exits) had been most unruly and unbecoming. Why the Dow Theory which hitherto has always allowed eschewing monster declines on this specific occasion failed to deliver a “safe” exit level with Chinese stocks. It is of little comfort to me to say, “Ok, US markets are different, they are more orderly and less volatile and if the past is a guide to the future we should expect the Dow Theory will continue to keep us safe."

I am very frank to you all: The pattern we saw with Chinese stocks is highly unusual. However, we would be ruined if two such “unusual” events happened in a row.

So, personally, what I have seen with Chinese stocks has been a wake-up call. I have not lost my faith in the Dow Theory, as I steadfastly continue to believe it is the best timing method by a long shot. However, we should be prepared for highly unusual scenarios.

This is why, as a reminder, it is very important to let Rhea’s page 77 book soak in (more about it here). Even though 90% of the primary bull/bear market signals are so-called “type 1” signals (secondary reaction, subsequent rally and violation of secondary reaction lows), occasionally the market starts to decline and refuses to stage even on one index a +3% rally. Here is when the “type 2” exit (the lows of the last completed, the previous one, secondary reaction) let us off the hook.


Thus, the “type 2” exit will offer us a technically sound (Dow Theory-based) exit point when the ordinary “type 1” exit fails to set up.

Furthermore, for those who think that the Chinese situation is not repeatable in US stocks, I’d like to make two observations:


Firstly, the Dow Theory is basically long term oriented. This results in few signals. Hence, we can say that for the last +115 year less than 80 signals have occurred (much less with the classical Dow Theory). While such a successful sample is reassuring because it spanned +115 years under very different economic conditions), it is far from being an exhaustive sample. I’d feel more comfortable if we had a +500 years sample with 400 signals. Therefore, albeit I remain committed to the Dow Theory, I am aware that more extreme scenarios are not to be ruled out. As traders say: “your worst drawdown is always ahead”.  Furthermore, please keep in mind that the US stock market has been very lenient with investors, as for over one century the US has been the economic leader of the world. Thus, our "sample" could be not so representative of extreme situations. There is a trick, though,  to suddenly expand your sample of trades by a factor of, at least x 26, which would be tantamount to ca. 2990 years. It is beyond the scope of this post to explain it, and, furthermore, I want to further research this issue. However, my preliminary results with such an expanded sample helped me to further cement my faith in the Dow Theory. All in all, if we manage to live 200 years, it is very likely that we will be confronted with more difficult price setups; however, the Dow Theory is very likely to fare well even under price structures not hitherto seen.  



Secondly, even with a +500 years track record, we are dealing with emotions, the economy and many variables we cannot control. Hence, a pattern (such as the one seen with Chinese stocks) that hitherto has not occurred with US stocks, could materialize in the future. Prices are free to set up as they please free of our intellectual constraints.

All in all, while I believe that the Dow Theory is great (and specially Schannep’s), I do also believe that we have to be armed with alternative exits when in the future we are confronted with a Chinese situation. The issue is how to get more premature exits when the price pattern, as interpreted by the Dow Theory, refuses to give us a “safe” exit.

What follows is the result of some days of instrospection. This is the roadmap as to "what to do" when the Dow Theory refuses to offer us an exit at decent levels:



a) Add third and even fourth index, in order to try to derive a Schannep DT signal. US stocks have the distinct advantage of having many indices. Please mind that we were dealing with just two indices when we were analyzing the Chinese stock market. Thus, if we see that we are approaching the “danger zone” (i.e. a loss exceeding 10-12%) and the indices we use refuse to set up for a primary bear market signal, we should start looking for greener pastures (i.e. including the NYSE, Nadaq, etc.) until we find a decent primary bear market setup. Please mind that this is very exceptional. Schannep is no friend of using the Nasdaq (pages 48-49 of his book "The Dow Theory for the 21st century"), and so am I. However, under exceptional circumstances (that is when by using the SP500, Industrials and Transports, we fail to get a decent primary bear market set up) we should try using other indices. With Chinese stocks you could be thinking in adding the CSI 300 Index (ETFs:  ASHR, PEK). By the way, this is one of the reasons why Schannep's Dow Theory is likely to continue outperforming (and cutting losses) better than the classical Dow Theory. With three indices is more unlikely that "Chinese" situations develop.



If no signal is derived with 3 or even 4 indices, then:

b) Look at what distance lies the 200 days moving average (if it lies 20% below current prices, then it is not a good alternative stop. However, if prices have been consolidating and the MA is at a reasonable distance (i.e. 12%), then this could be our alternative (which we accept reluctantly) stop. Readers of this blog know that I am no friend of moving averages as timing indicators, as the Dow Theory is beyond one shade of doubt clearly superior, as explained here.

However, we are now dealing with exceptional measures, and, hence, on the extremely rare occasions when the Dow Theory fails to facilitate an exit at a reasonable level, we have to look for alternatives. The 200-day  MA is an alternative.

In addition (or in lieu of) to the MA stop, we should start accepting alternative technical stops as those three listed below under letters “c”, “d” and “e”.

c) Minor lows that did not qualify for a secondary reaction (due to lack of time and/or extent), apparent on the chart, which are within a reasonable distance (i.e. no too near, at, i.e. -4% below the peak, but not too far away at -20%). After all, according to Rhea, the violation of any low is certainly not bullish.

d) Is the violation of a significant trend line at a reasonable level (i.e. -10% below the top) and not too near (i.e. -4% below the top).

e) Is there a head and shoulders pattern, which allows for an exit at a reasonable level?

f) Honor a percentage fixed (i.e. -16%) stop. Bear in mind though that what has worked for the US stock market (and there is no guarantee that an empirically-based stop will continue working in the future even for US stocks) may not necessarily work for the Chinese stock market which tends to be more volatile, and hence might require a more ample stop. 

To further cement your confidence in the alternative exits listed above, pay special attention to non-confirmation or divergences as well. The current break down of Chinse stocks was preceded by a non-confirmation, and I alerted my readers here. If there was a non-confirmed top, the odds for a break would be higher. Under such a scenario, and when no “normal” Dow Theory stop is at a reasonable distance, then be willing to use less orthodox ways of exiting.

Of course, it’s up to each one to decide what to do. I can only provide a small blueprint.

Sincerely,
The Dow Theorist



No comments:

Post a Comment