“Cutting immediately to the chase, we continue to believe that the U.S. equity market is in a late-stage top formation of the third speculative bubble in 15 years. On the basis of measures best correlated with actual subsequent S&P 500 total returns across history, equity valuations remain obscene. We fully expect a loss in the S&P 500 in the range of 40-55% over the completion of this cycle; an outcome that would be wholly run-of-the-mill given present market conditions, and would not even bring reliable measures of valuation materially below their longer-term historical norms.” – John Hussman, PhD., Weekly Market Commentary from Hussman Funds at…
http://www.hussmanfunds.com/wmc/wmc151116.htm
2-to1 ODDS THAT THE BULL MARKET HAS ENDED (Navigate the Stock Market)
I think the odds that we are now in a Bear market are about 66%, or 2 to 1. Here’s why…
If one visits any of the discussion boards on stocks, there are always strongly held opinions about a bull or bear market underway. Rarely do posters place a probability on their opinions. One method that could be used would be to assign probabilities to various stats and use Bayesian methods to estimate conditional probability. Thomas Bayes’ 18th century method of analysis is widely used today. To use Bayesian methods, we need three estimates regarding the probability of the subject under consideration. In this case, “Are we in a Bear market?” First, one must have the “Prior” probability.
The odds of a Bear market, absent any other information, are about 30%. (Stock Market Logic, by Norman Fosback. Simply put, the stock market is in a bear about 30% of the time.) So the initial "prior" estimate that we are now in a bear market is 30%. To use Bayes Theorem, we need more information and the two more probabilities. We need the probability associated with the new information conditional on the idea that a Bear market is underway; and the probability that the new information is not conditioned on the Bear market thesis. Let’s look at ISM and FED data that John Hussman, PhD, referenced last week in his Market Commentary.
ISM and FED Manufacturing data has been down sharply 3-times since 1995 and that preceded 2-Bear Markets and one sharp correction. A simple view of the world is that the with ISM and FED Manufacturing numbers sharply falling a crash will occur 2 out of 3 times, or 66%. So that is the probability we might assign conditioned on the thesis being true that the ISM and FED data support a Bear Market thesis.
Now there could be other factors that might prevent a bear market; the manufacturing data could improve or the strong ISM Services data might forestall a downturn. Let's assign 50% as the odds that No bear market will be associated with ISM and Fed Manufacturing data, i.e., the data is not conditioned on a Bear Market. (Note that the probabilities don’t need to add to 100%, because the result is not simply one option or the other. There can be other factors that play into the null hypothesis. This is a strength of Bayesian analysis.)
Those new probabilities are inserted into the equation (bottom line in the table) along with the prior probability of a Bull Market Top / Bear Market Start and the new probability is calculated in the Table below:
One can see that the new probability has climbed to 36%
that the Bull Market is over and a Bear Market has started.The prior probability of 30% is now been superseded and the new "prior" is 36%. We can now consider further new information such as earnings and revenues. Company earnings and revenues have declined for the last 2-quarters. Earnings have declined in the past without a bear market, but it does seem to make it more likely that we could be in a bear market; therefore let's assign a probability of 55% that we are in a bear market. The converse probability is hard to estimate so let's say the odds that earnings will improve is 50%.
The prior probability of 36% is now been superseded and
the new "prior" is 38%. We can now consider further new information.The S&P 500 chart looks like it is a top. Charts have looked like this in 2000, 2007, 2011 with crashes in 2000 and 2007. It sure looks like a top so let’s say the odds are 2 to 1 a crash will occur based on the charts. There was no crash in 2011 (only a bad correction) so there is a 1/3 chance of NO crash.
Based on the information examined so far, the odds of a
Bear Market underway have climbed from an initial 30% to better than even 55%.
Further, the new-high/new-low data looks bad. In fact only 2.3% of stocks made
new highs at the most recent high on the S&P 500. At the 1929 top only 2% of stocks made new-highs. In 2000, 11% of stocks made new-highs at the S&P 500 top. The probability that a Bear Market has started conditional on only 2.3% of stocks making new highs at the S&P 500 top can be roughly estimated as follows:
Let's consider data after the new-high peak in Jan 2013 when the %-new-highs was 23%. That was the peak of new-highs considering all S&P tops going back to the 2009 bear-market low. We can assume that if 23% of stocks are making new highs the odds of a bear Market are near zero. If no stocks are making new highs at an S&P 500 new high, we might assume that the odds of a bear market are close to 100%.
Using a straight line assumption, the odds of a bear, given 2.5% new-highs = 100-(2.5/23)*100=90%. (An asymptotic assumption might be better than a straight line, but given the rough nature of this analysis, maybe not. I have also put a higher probability on the null-hypothesis (below) and that counters the high probability for a crash.)
There could be other mitigating factors (such as continued investor exuberance) that would indicate that 2.3%-new-highs do not indicate a final high for the S&P 500. Excluding the latest new high on the S&P 500, during 2015 there were 9-new S&P 500 highs with an avg. % of stocks making new highs of 4.5%. The lowest %-new-highs were 3.5%, 2.7% & 3.1%. None of those low readings were the top because in each case the market went on to make higher-highs. So there is a reasonable probability that the S&P has not entered a Bear market even with only 2.3% new-highs at the prior top.
Let’s assign a probability of 50% that there is not a Bull Market top associated with 2.3% of stocks making new highs. Calculating the new probability with all of the new information shows the following:
Thus, the probability that a Bear Market is underway is
estimated to be 69% - better than 2 to 1 odds. (Incidentally, the order in
which the analysis is made makes no difference to the overall result.)While this analysis presents my opinion in a probabilistic approach, some would say that view is skewed by not including good news. For example, analysts presently predict that earnings will improve in Q4 of 2015 and turn positive in Q1 of 2016. Economists predict GDP will improve in 2016. That good news could be added to the analysis and potentially reduce the probability that we are now in a Bear Market. By ignoring it, I am tacitly assigning a 50-50 probability that the analysts and economists are correct.
Why give analysts and economists so little credit? “Consider what happened in November 2007…one month before the Great Recession began…Economists in the Survey of Professional Forecasters…expected the economy to grow at just slightly below 2.4% in 2008….GDP actually shrank by 3.3%” – Nate Silver, The Signal and the Noise…why so many predictions fail – but some don’t.
The probability above is fluid and would be influenced positively, or negatively, by news to come. I am sure that statisticians would find fault with my probability estimates and possibly the methodology; but after all, this is the stock market and there is so much data that most prognosticators are reduced to guessing. This analysis is simply an attempt to make a more educated guess.
MARKET REPORT / ANALYSIS
-Monday, the S&P 500 was up about 1.5% to 2053 at the close.
-VIX was dropped about 10% to 18.16.
-The yield on the 10-year Treasury slipped to 2.27.
A strong snap-back rally is typical in downturns; Monday the Index snapped up. Monday was another statistically significant day and that means simply that the price-volume move exceeded my statistical parameters and, in about 60% of the time, that leads to a down-day the next day.
The S&P 500 Index is still 0.5% below the 200-dMA, so that potential level of support is now resistance for the S&P 500 to move higher.
My longer term guess is that the market moves generally down for a bit more. The 50-dMA on the S&P 500 is 2008. A 50% down retracement would put the market at about 1990. The chart looks like an important level is around 1930-1940. All of those levels should be watched for a possible buy signal.
MARKET INTERNALS (NYSE DATA)
The 10-day moving average of the percentage of stocks advancing (NYSE) fell to 43.4% Monday vs. 44.3% Friday. (A number below 50% is usually BAD news for the markets. On a longer term, the 150-day moving average of advancing stocks was up slightly to 49%. (Lowry Research considers the 150-day advance decline time frame to be a critical measure of longer-term, market health.) A value below 50% indicates a down trend.
The McClellan Oscillator (a Breadth measure) remained negative Monday.
New-lows outpaced New-highs Monday. The spread (new-highs minus new-lows) was minus-111. (It was -195 Friday.) The 10-day moving average of the change in spread was minus-18 Monday. In other words, over the last 10-days, on average; the spread has decreased by 18 each day. The internals remained negative on the markets.
NTSM
Monday, the NTSM long term indicator was HOLD. The Price indicator is positive. Volume, VIX and Sentiment are neutral.
MY INVESTED STOCK POSITION:
TSP (RETIREMENT ACCOUNT – GOV EMPLOYEES) ALLOCATIONAll cash: G-Fund (Cash, risk-free yielding 2.1% over the last 12-months): 100%
I made a rather impulsive decision. For my reasons (or lack of reason) see “My Invested Stock Position” in my prior blog at...
http://navigatethestockmarket.blogspot.com/2015/11/factset-earnings-cass-freight-index.html
My current thinking on the market is explained above in the paragraph, “2-to1 ODDS THAT A TOP IS IN AND THE BULL MARKET HAS ENDED”. Be aware: My recent success in calling tops is not good. If I am wrong, I’ll move back in.
