Stock market trading based on 6 month stock market forecasts - So what?

Why pay any attention to the stock market forecasts from this forecasting model? It looks like a buy/sell strategy may significantly boost investing returns.

In my last post I introduced a stock market trading strategy based on my 6 month market forecasts. It turns out that the forecasts consistently lead the market by a couple of months, so actual trading works best not by following the most current forecast, but by using a lagged forecast that weights forecasts from a few prior months.  The buy/sell periods called by that strategy are shown in the plot at the end of this piece. That plot shows the strategy produced just a handful of handful of buy/sell points in the period since mid-2007 when I started posting forecasts for my models.

So what?  A few correct buy/sell decisions can have a huge impact on investing returns. In particular, the forecasts of 2008 through early 2009 screamed of likely 6 month market losses of  17%  or more, months before the crash took place.  They turned out to be right. They were worth paying attention to.

A back-test applying the buy/sell strategy to an S&P 500 index fund is shown below running from mid-2007 to the present.  In the stock crash of 2007-2009 the S&P 500 took a big hit, but overall for the period the average has had a 4% annualized rate of growth. The strategy using the 6 month stock market forecasts, however, had an annualized growth rate of 14%, roughly 3 times better.

(Click on image to enlarge.)

The big market call of major likely declines made for most of the difference in performance.  Since the bottom of the financial meltdown the strategy has had only a few brief sell periods. So, for mid 2009 to the present the S&P 500 index has had a great rise and has only slightly under-performed the strategy.

The chart below shows the buy/sell calls produced by a strategy based on using a weighted average of 6-month market forecasts.

(Click on image to enlarge.)

Stock market trading based on 6 month stock market forecasts

I am adding a stock market trading strategy to my monthly blog posts, a Buy-Sell-Hold  model for the U.S. stock market based on a weighted average of my forecasts from prior months.

From the start, there has been a problem making use of my six month stock market forecasts:  what's going to happen in the stock market six months from now has very little correlation with what the market is going to do in the next few days and weeks. (Just because I know it is going to get hot next summer, that's not much help in knowing whether I need to wear a coat outside in mid-November.)

The solution I have come up with is to develop a stock market trading strategy based on a statistically weighted average of my 6 month stock market forecasts from prior months, not the current prediction.

Below there is a plot of the Buy/Sell calls from the model for the S&P 500 index over the past several years.  Most of the time the strategy yields a clear Buy/Sell opinion, but there are a few spans of time when the model had no real opinion.

(Click on image to enlarge.)

Stock Market Forecast November, 2015 through April, 2016: Buoyant

Stock market prospects for the next 6 months are strong. There is about a 90% probability that the market will go up.

The stock market forecasting models documented here are based on the premise that the broad U.S. stock market tends to do approximately what it usually tends to do. Over a forecasting horizon of six months the market will respond to the same real economic factors that have moved prices over the past several decades. Shocking, right?

In late August through September the stock market was panicked by fears of things that don't have large real dollars and cents impact on the U.S. economy (financial uncertainty in China and Greece along with weak world oil prices).  My models had expected typical seasonal market weakness, but not the extra downdraft of international finance fears.  Then, just like the not-at-all-real goblins of Halloween, in October those extreme fears fell away and the market staged a sharp recovery.

Over the long haul my experience-based stock market forecasting results tend to be pretty good. Or, at least they have proven to be fairly accurate since 2007 when I started publishing forecasts in real time.

Here is my 6-month forecast:
U.S. Stock Market Forecast (Value Line Arithmetic Index):
Probable stock market gain 11/1/2015 to 5/1/2016: 11% (Avg. 6 mo. gain since 1984: 4.8%)
Probability of at least breaking even : ~90%  (Average for all months since 1984: 73%

Here is the track record of my 6 month stock market forecasts since 2007:

(Click on image to enlarge.)

Stock Market Forecast October, 2015 through March, 2016: Bouncing back

Last month the title for my stock market forecasting entry was also "Bouncing Back" and that was certainly not the case for the lousy stock market September we just got through!

But, this blog and my stock market prediction models focus on the coming six months, not the next few weeks.  Take a look at the stock market performance versus prediction graph below.  The model had been expecting seasonal weakness for the late summer and that is what occurred, just worse than forecast.   If my macroeconomic model is to be believed, it says that the market is over reacting to international news and things should start to get better over the next few months.

(Click on image to enlarge.)

So here is my 6-month forecast:
U.S. Stock Market Forecast (Value Line Arithmetic Index):
Probable stock market gain 10/1/2015 to 4/1/2016: 10% (Avg. 6 mo. gain since 1984: 4.8%)
Probability of at least breaking even : 85%  to 90%  (Average for all months since 1984: 73%

Stock Market Forecast September, 2015 through February, 2016: Bouncing back

My econometric stock market models are predicting a significant stock market recovery of about 10% over the coming 6 fall and winter months of 2015-2016. That bounce is despite, or more accurately, because my models had not expected the market to tank so badly during August.

Huh?

My forecasting models totally missed expecting the mid-August market panic / correction. The prediction models had been expecting a sub-par market over the summer months, reflecting the statistical fact that stock markets tend to be weak over the summer. But, no way did the models foresee the abrupt market correction that actually, and painfully, happened.

So, how is that useful information?  The answer is that the fears that moved the market (high valuations, economic weakness in China, low commodity prices, and a weak oil market) are not the basic economic forces that typically have a lasting and remarkably predictable longer term impact on U.S. stock prices. In short, the models say that the market over-reacted to bad news that really is not that important to U.S. stock prices. If history is a guide, the positive economic fundamentals will regain their importance fairly soon and the market will recover.

Morningstar.com agrees. Their Market Fair Value graph, based on net present value calculations today says that the overall market is 7% undervalued. Stocks are on a 7% off sale! (I remain convinced that this Morningstar graph is the most useful stock market indicator available for free on the web.)

So here is my 6-month forecast:
U.S. Stock Market Forecast (Value Line Arithmetic Index):
Probable stock market gain 9/1/2015 to 3/1/2015: 10% (Avg. 6 mo. gain since 1984: 4.8%)
Probability of at least breaking even : 85%  to 90%  (Average for all months since 1984: 73%

Stock Market Forecast August, 2015 through February, 2016: A bit better

I love a boring summer stock market.  Nothing much seems to be happening.  We in the U.S. are in the middle of the typically weak summer period. So,looking forward towards winter when the market typically had recovered, 6 month stock market performance will probably be a little better than average.

Here is what my econometric models are predicting:
U.S. Stock Market Forecast (Value Line Arithmetic Index):
Probable stock market gain 8/1/2015 to 2/1/2015: 6% (Avg. 6 mo. gain since 1984: 4.8%)
Probability of at least breaking even : ~ 80%   (Average for all months since 1984: 73%

So, why is a boring stock market wonderful?  Because the market is unlikely to crash really soon. Some things in life are very simple.

If all things happen as they 'typically' do in the economy, my personal bet is that the market will have its next serious upset in a couple/few years. (My mathematical models have no opinion on this -- they only look 6 months ahead.)

Here is how the the predictions of my forecasting models and actual performance of the stock market have landed over the past 8 years:

(Click on image to enlarge.)

As shown in the chart above, the U.S. stock market (measured by the ValueLine Arithmetic Index) has done pretty much what my models had expected for the past year or so. Actually, my forecasts have matched the market unusually well.  Since the models are based directly on what the market usually does in response to macro economic factors, this probably means that no really major strange outside factors are driving the market.

The great protective shield of historically low interest rates that has been maintained by the U,S, Federal Reserve Bank has made this huge bull market happen.  It is pretty easy for businesses to make money when they can borrow money for next to nothing. In a few years the current odds are that the U.S. economy will have become frothy. That's when the Fed will jack rates up rapidly and 'take the punch bowl away from the party'.  But, that's what is likely to happen in a few years -- not now.  Enjoy summer!

Stock Market Forecast July, 2015 through January, 2016: Average

The first half of 2015 saw the U.S. stock market bounce around, but end up just about where it was at the start of the year.  Performance had been very close to our models' projections until this past week when concerns over the Greek debt crisis caused markets world wide to stumble.

Our models predict that the second half should start to get better toward the end of the year, but still fall a bit below normal performance.  As far as the model is concerned, it is anticipating normal summer weakness in the course of a maturing -- and slow growing -- bull market. Not much reason to either sell or buy.

U.S. Stock Market Forecast (Value Line Arithmetic Index):
Probable stock market gain 7/1/2015 to 1/1/2015: 4% (Avg. 6 mo. gain since 1984: 4.8%)
Probability of at least breaking even : ~ 70%   (Average for all months since 1984: 73%

(Click on image to enlarge.)

Looking at the graph above, for the past several months our models have predicted weakening market results.  The market in real time appears to be flagging faster than expected.  For the 6-month period just completed, our forecast have been for roughly a 5% gain.  As of last week the market had logged a 4% gain.  But, thanks to fears brought on by Greek sovereign default, the actual  six month result was no gain or loss.  Looking at prior years on the chart, when the market falls faster than predicted it also tends to stay weaker then anticipated for several months.  Because of the uncertainty of pending debt default in Greece and Puerto Rico, it wouldn't be surprising to see more market weakness in the next few weeks. A steep enough tumble might open up a buying opportunity.