Tuesday, December 6, 2016

Dow Theory Update for December 6: More on the primary bull market signal

Trends unchanged for gold, silver and their miners: that is bearish

On November 21st, 2016 a primary bull market was signaled, as was reported here

Let’s further delve into the signal. First of all, let’s take a look at the charts

The highs of the last primary bull market (red horizontal lines) were bettered by the three indices: New primary bull market
After the primary bear market signal of November 2rd, 2016 (signal which has not heeded by Jack and Bart Schannep, as explained here), US stock indices rallied with no respite until the last recorded primary bull  market closing highs (red horizontal lines and blue arrows) were jointly bettered by the Industrials, Transports and S&P 500. The Transports (middle graph) were the first to better their last recorded primary bull market highs (11/7). The Industrials did so on November 10th, and eventually the S&P 500 confirmed on November 21rd, thus being signaled a primary bull market signal. Please mind that as per Rhea’s and Schannep’s Dow Theory the breaking out of the last primary bull market closing highs constitutes a legitimate primary bull market signal. While this is not the most usual primary bull market signal (the “usual” one entails the development of a secondary reaction against the primary bear market, subsequent pullback and breaking out of the secondary reaction highs), it is a fully valid one.

All in all since November 21st, 2016 the primary trend is bullish. Furthermore, as per the classical Dow Theory (and according to each possible interpretations thereof), the primary trend is bullish as well.

As to the Dow Theory when applied to weekly bars, while I still didn’t have the time to write an in-depth post, the trend has turned bullish, which is supposed to be tailwind for the current primary bull market signal (as determined by the daily bars). In other words, the trend in the higher time frame (weekly bars) is supportive and in gear with the trend in the lower time frame (daily bars).

So overall, technically at least, the trend is very bullish. I see, though, two clouds on the horizon (which nonetheless, I disregard, as I am strictly a technical trader). Please mind that “clear skies” in real time trading are a very rare occurrence. This is why it is so difficult to be a rule-based trader. One thing is back testing, another quite different thing is to trade in real time and stick to the rules no matter how many “clouds” we see on the horizon:

The first issue that bothers me is the “narrow ranging” which has been plaguing the stock market since 2010. The second issue is the age of the current cyclical bull market. I will further elaborate on both “clouds”.

Since 2010 the market has been experiencing some sort of “fibrillation”. In other words, further declines following primary bear market signals are short lived and shallow, whereas further advances following primary bull market signals lack enough magnitude before being reversed by the next primary bear market signal. I have already written about this unpleasant market environment here.

I have tabulated all trades taken as per Schannep’s Dow Theory since 1953. I calculated, among other items, the total amplitude of an upswing (that is from primary bear market low to primary bull market top), and the amount of the further advance until the top following a primary bull market signal. These are two key figures. The only way for any Dow Theory flavor to thrive is to have swings (from bottom to top) of sufficient magnitude so that if we deduct the percentage lost when entering (distance between entry price and bear market bottom) and exiting (distance between market top and exit price) we still have a healthy remainder: this remainder is the “meat” or our takeaway home from each Dow Theory swing. We know that we shed ca. 7-8% when entering and exiting. Thus, we can say that roughly we stand to lose ca. 15% of each upswing. If the upswing has a magnitude of let’s say 40% we will stand to make ca. 25% on the trade. If the upswing has a magnitude of just 15%, we will likely break even. A narrower range will entail outright losses. Please mind that this is not carved on stone. These are just rough calculations which, though, are well grounded in the historical record.

We do not only need a sufficiently ample upswing. We also need that the further advance following the primary bull market signal has enough “oomph” and is not aborted soon after the primary bull market signal.

From 1953 until 2009, the average upswing (from bear market low to bull market low) amounted to 54.15% (median 39.59%). Impressive, isn’t it?. Let’s take a 40% figure (close to the median). If we subtract 15% (ca. 7-8% lost when entering and exiting) we have a likely gain of ca. 25% per trade. By the way, this boe estimate quite accurately reflects the historical Dow Theory record: 25% made in ca. 1 ½ years (roughly estimate of the average trade duration) translates into a ca. 16% p.a. performance which is not far from the historical Schannep’s Dow Theory record until 2009.

Well, now fasten your seat belts: From 2010 to date, the average upswing amounted to just 25.88% (median 23.33%). In other words, the average upswing has been ca. cut in half. In my opinion this is clearly “fibrillation” and headwind for the Dow Theory. Noise has dramatically increased (“noise” being the lesser amplitude of upswings versus the amount of signals).

And what about the subsequent advance following a primary bull market signal. In other words, after buying following a primary bull market signal, is it very likely that the market will immediately reverse and hence we will make a loss or realize meager gains?

From 1953 to 2009 the further advance following a primary bull market signal was 37.98% (median 27.03%). In other words, if we subtract ca. 7-8% (the exit tends to be ca. 7-8% below the primary bull market top), we stand to make ca. 20% (more if we take the average instead of the median) on each trade.

However, from 2010 to date, the further advance following a primary bull market signal has been maimed. The further advance following a primary bull market signal was a meager 12.16% (median 7.6%), hence leaving very little “meat” available to profit from. If we deduct ca. 7-8% from the top, we are just close to break even. No loss, but, of course, very little gain. 

Note: All calculations have been made following the S&P 500; calculations are similar but not identical using the Industrials or the Transports.

If we couple lack of follow through following a primary bull market signal with an equally maimed further decline following primary bear market signals (which is the main driver of outperformance versus buy and hold, the larger the decline following the primary bear market signal, the better for the Dow Theory, as I will prove with numbers in a future post), we have a quite dismal picture. When getting out (primary bear market signal), the decline stops and we are forced to re-buy (next primary bull market signal) at normally higher prices, and when getting in (buying) the market is close to making a top and hence we can hardly build up profits.

As I wrote recently:

Let’s be clear about one thing: If I were to believe that from now on future primary bear market signals would be met with scant follow through, then I ‘d be better deserting the Dow Theory for good (at least Schannep’s which in the only “flavor” with reliable statistics and hard and fast rules, as the “classical” sadly is plagued with many subjective interpretations). In other words, if further declines following a primary bear market signal as determined by Schannep's Dow Theory were an oddity of the past (“oddity” which has persisted since 1960 even in recent times, albeit to a more mitigated extent), we should declare Schannep’s Dow Theory as dead, as the main source of outperformance versus buy and hold (further declines following the primary bear market signal) would be cut short.

The very same remarks I made concerning the lack of follow through after a primary bear market signal are fully applicable to the lack of further advance following a primary bull market signal. Further declines after the bear market signal are our main source of outperformance whereas further advances after the bull market signal are our main source of raw performance. Absent the two the Dow Theory will underperform and have meager returns. Please re-read the previous sentence. If we just had lack of further decline following primary bear market signals but further advances following primary bull market signals were sound, we would be underperforming but, at the same time, we would be having solid performance (i.e. buy and hold 20% p.a. and the Dow Theory 15% p.a.) which is a quite comfortable situation, given that sooner or later the Dow Theory will save us from a crash (or big decline).

On the other hand big declines following a primary bear market signal coupled with scant follow through following a primary bull market signal would imply Dow Theory outperformance versus buy and hold, but, at the same time, a likely negative performance (i.e. buy and hold, -20% versus Dow Theory -5%).

Thus, while it is unpleasant to underperform, it should give us solace to have, as we have, a keen and deep understanding of the nature of the underperformance.

My personal take is that Schannep’s Dow Theory (and the Rhea/classical as well but with more likelihood of higher drawdowns) is very solid. Why am I so sure? As an appetizer, please read here and here. I am convinced that Schannep’s Dow Theory (and of course the “Rhea/classical”) is not dead. The current ranging environment will sooner or later end. Rough patches plague all trading systems (by the way all trend followers are suffering, not just the Dow Theory) and we have to learn to live with them. The only way to live through the rough patch is to deeply understand the trading system. And this is hard work. No way around it. 

The second issue that bothers me is the age of the current cyclical bull market. The current cyclical bull market (more about cyclical bull markets here) is already very old. However, in real time it is very difficult to predict its demise (as in real time there is almost a distinct 50% probability of it continuing for some more months and percentage points). Furthermore, a last "gasp" of +20% is more than enough to close the current trade with profits, so one never knows. Hence, rather than relying on the statistical record of past cyclical bull markets, I am more confident on the solidity of the Dow Theory to ride different market environments. All in all, the current cyclical bull market got started in 2011, and hence it is an “old” one which implies that the odds a cyclical bear market (more than -16% decline on both the S&P 500 and Industrials as per Schannep’s definition) are high.

The Dow Theorist

Thursday, November 24, 2016

Dow Theory Update for November 24th: More about the primary bull market signal for US Stocks

How Jack and Bart Schannep avoided the primary bear market trap

In my last post I briefly mentioned that the primary trend had turned bullish as per Schannep’s Dow Theory. As explained, the primary trend was already bullish according to the “Rhea/Classical” Dow Theory, as was explained and summarized here

Before continuing, I’d like to make clear that Jack and Bart Schannep themselves did not heed the primary bear market signal of November 3rd, 2016, and hence they recommended their subscribers to remain fully invested. Even though they identified the signal on the charts, they decided not to act upon it. This decision was not whimsical and was really well grounded: Jack and Bart Schannep realized that past primary bear market signals lacked follow through. They realized that if one had demanded an additional 3% further decline following the primary bear market signal on both the S&P 500 and the Industrials before acting most recent whipsaws would have been avoided. 

The chart below displays both the primary bear market signal of November 3rd (not acted upon by Schannep) and the November 21st primary bear market signal (to be acted upon by those that exited the market on November 3rd):

Whipsaw avoided by those following Schannep's temporary new rule

Jack and Bart Schannep are aware that this “new” rule (demanding and additional 3% further decline after the primary bear market signal) is just temporary and meant to be applied only during this bothersome “fibrillation” spell. The one million question is: When will the fibrillation spell end? This is a tough judgment call. If “fibrillation” is to last for some more trades ahead and the behavior of US stocks indices close resembles what we have seen since 2010, then Jack and Bart Schannep are fully right. If either the narrow ranging ends soon or the narrow ranging continues but not exactly as it was until now (i.e. following the primary bear market signal, the SP500 and Industrials decline, i.e. a further 3.2% thus flashing a “sell” under the new rule just to see the decline end immediately thereafter) then the new rule would prove to be detrimental to performance.

Jack and Bart Schannep feel that the current market environment may last for a while, hence warranting the temporary “tweak” of the rules. This is, of course, a decision not based on technical analysis and involves gut feeling, something which is not measurable.

On the other hand, as it was explained to me in an email dated November 4th, it seems that Jack and Bart Schannep will be more likely to use the new “3% additional decline” rule when the original sell signal (not to be heeded) is flashed at a narrow distance from the last recorded primary bull market highs (as was the case with the bear market signal of November 3rd). This makes sense. It is not the same to require an additional 3% decline following a bear market signal that has been signaled close to the top (i.e. 4%) than a bear market signal flashed at 10% below the last recorded bull market highs. So it seems that the application of the “tweaking” will be nuanced (once again involving a further judgment call)

I know that goal number one in the markets is to make money (or avoiding being whipsawed which is the same), and hence one should not get too dogmatic about being reluctant to bend the rules if need be. Best traders have always had the right instinct to break or bend the rules when things change. The issue is determining whether things have really changed or it is merely a rough patch of trades which is about to end. This is what separates the best traders (i.e. Jack and Bart Schannep) from the rest of the pack.

My trader background is that of a strict rule-based trader. Hence, in my short term trading I abide to my rules. Period. No exception. Experience showed me (at least for me of course) not to tweak rules and to admit that the market will not always oblige.

Maybe because in the first place I am not a legend, and I lack the instincts of top traders I am forced to settle with strict rule-based trading if I am to survive in the tough world of investing and trading. Being my gut feeling not so gut, I need to rely on clearly defined set of rules.

Accordingly, I decided to heed the primary bear market signal as “it was” without second guessing. Maybe the best explanation for not having applied the “tweak” to the rules suggested by the Schannpe’s is that I lack the flexibility and acumen of the best traders. However, and since I do eat my own cooking, I’d like to summarily state the reasons that prompted me to strictly stick to the rules:

1) In general, the less rules the better (less degrees of freedom). The more rules the higher the risk of so-called over fitting, which means that the added rule may work for a while and cease abruptly to work in the future.

2) The new rule of demanding an additional decline of 3% (in both the SP 500 and Industrials) to act on the primary bear market signal is tantamount to eliminating the very source of outperformance of Schannep’s Dow Theory versus the “Rhea/Classical” Dow Theory. Of course, if the new rule is merely “temporary” and Jack and Bart Schannep apply it exactly when it needs to be applied, there is no risk of eliminating outperformance. My big issue is that in real time I cannot predict whether the sell signal is a false one (whipsaw) or whether a big decline will ensue. However, very good traders (and Schannep belongs to this league) may “sense” when is the right time to use or disregard the new rule.

3) If US stock indices where to go back to their “normal” behavior up to 2010 the average and largest losing trade is going to go up as a consequence of the new rule. By the same token, the largest drawdown is likely to increase.

4) The likelihood of being caught in a big crash is going to increase. The more time (and the more decline), it takes for us to exit, the higher the risk of being caught unawares by a crash. Of course, if we lose big one time (i.e. 30%) due to the new rule, and ten times we make extra 5% due to the avoidance of whipsaws, the new rule is worth (provided one’s drawdown tolerance accepts occasional huge drawdowns).  In real time nobody knows.

5) Even though in recent times declines have been more modest, there continues to exist further decline following Schannep’s Dow Theory primary bear market signals. Thus, following the exit of 12/11/2015 there was a further 9.11% decline.

6) Demanding an extra -3 % decline following the primary bear market signal, implies a reduction of the further decline, which is detrimental to performance. The tradeoff is as follows: In order to avoid whipsaws, one settles with larger losers (because one gets out later) and smaller winners (as the subsequent decline has been reduced and hence one is reducing the source of outperformance. Of course, the number of trades and specially, losing trades is reduced under the new rule. I concede. 

7) Basically, exiting at ca. -3% below the actual signal, implies giving up 3% performance if markets where to return to the pattern seen until recently. Eventually, we have to make a judgment call: Are narrow ranges being the norm for the next few years? Are US indices going to have ample swings soon? I don’t know, but I tend to lend credence to the historical record of swings. I feel we are living an aberration, albeit painful.

Of course, the above mentioned line of thought is worth nothing against a good trader with a well fine-tuned gut feeling.

Thus, for those following Schannep’s advice, the current primary bull market signal of November 21st was immaterial, as they had remained fully invested.

One thing is clear: These narrow ranging conditions are putting trend followers to the test. It is easier to write about underperformance than living through it.

When almost one year ago I decided to start the “Dow Theory stress saga”, I wanted to imagine hitherto unseen difficult scenarios for the Dow Theory. Well, markets have obliged, and the last year has been really challenging. It’s been a real stress-test. Not only for the Dow Theory but, in general, for all trend followers, as breakouts and breakdowns lack follow thru.

Thus, until the recent past, we could be very confident that following a primary bear market signal, stocks would further decline. By the same token, following a primary bull market signal, we could count that in most instances there were to be significant further advances. However, this has not been the case in the last few years which resulted in killing the very source of outperformance (the advantage of trend following versus buy and hold is to “cut losses short”, namely that following a bear market signal, the markets continue falling whereas one keeps powder dry). It is as simple as that. If no further decline following a primary bear market signal, no outperformance. Period. 

The one million question now is: Will future primary bear market signals be met with further declines (and hence the Dow Theory will continue to work after a temporary rough patch)?

Or put it alternatively, will US stocks continue to tease us and stop declining after primary bear market signals?

Here we must take a stand. If we believe that the market has changed its nature forever, we should make amendments to the Dow Theory (which for me is entering uncharted waters). If we believe that we are merely living through a rough patch (no matter how painful it is) then we should with discipline follow the Dow Theory as is. A third alternative, fully respectable, is that followed by Jack and Bart Schannep: On a temporary basis (until they deem that market conditions have changed) they will demand an additional decline of 3% in order to act on a primary bear market signal.

I feel this post was important to summarize where we stand now. 

The next post will focus on the analysis of the primary bull market signal.

The Dow Theorist