Preface: Explaining our market timing models
We maintain several market timing models, each with differing time horizons. The "Ultimate Market Timing Model" is a long-term market timing model based on the research outlined in our post, Building the ultimate market timing model. This model tends to generate only a handful of signals each decade.

The Trend Model is an asset allocation model which applies trend following principles based on the inputs of global stock and commodity price. This model has a shorter time horizon and tends to turn over about 4-6 times a year. In essence, it seeks to answer the question, "Is the trend in the global economy expansion (bullish) or contraction (bearish)?"

My inner trader uses the trading component of the Trend Model to look for changes in the direction of the main Trend Model signal. A bullish Trend Model signal that gets less bullish is a trading "sell" signal. Conversely, a bearish Trend Model signal that gets less bearish is a trading "buy" signal. The history of actual out-of-sample (not backtested) signals of the trading model are shown by the arrows in the chart below. Past trading of the trading model has shown turnover rates of about 200% per month.

The latest signals of each model are as follows:
  • Ultimate market timing model: Buy equities*
  • Trend Model signal: Neutral*
  • Trading model: Bearish*
* The performance chart and model readings have been delayed by a week out of respect to our paying subscribers.

Update schedule: I generally update model readings on my site on weekends and tweet mid-week observations at @humblestudent. Subscribers will also receive email notices of any changes in my trading portfolio.

A lukewarm buy signal
Mark Hulbert recently updated the market forecast of former Value Line research director Sam Eisenstadt. Eisenstadt has had a remarkable record of forecasting equity returns, according to Hulbert.
The reason to take this projection seriously is Eisenstadt’s track record. Consider a statistic known as the r-squared, which measures the degree to which one data series predicts or explains another. If the first series perfectly predicted the second, the r-squared would be 1.0; if the first series had absolutely no predictive ability the r-squared would be zero.

For the data plotted in the chart below, the r-squared is 0.31, which is significant at the 95% confidence level that statisticians often use when determining if a pattern is genuine. Though you might be disappointed that this r-squared isn’t higher, you should know that most of the models that get attention on Wall Street and in the financial press have r-squareds that are far lower—if they’re not actually zero.

Eisenstadt`s latest six month SPX forecast is 2620 to 2640, which represents a gain of about 3%. The chart shows Eisenstadt`s forecasts, as documented by Hulbert, since 2013. The forecast levels are shown in blue, with the actual below (black if the market beat his target, red if it missed).

Hulbert wrote that Eisenstadt has two critical inputs to his forecast, both of which are mildly bullish:
Eisenstadt constructs his model to include all factors he has found to have an ability to project the stock market’s subsequent six-month return. Though his model is proprietary, Eisenstadt has told me that two of the more important inputs are low interest-rates and market momentum. Both factors are mildly positive right now.
No forecast is complete without some understanding of the risks to the forecast. When I peek under the hood of the Eisenstadt's model, interest rates and momentum represent sources of both risk and opportunity to the market. Investors will have to judge for themselves whether a potential 6-month gain of 3% is worth the risk.

The full post can be found at our new site here.