Subject: Re: A really new strategy (maybe)
Thanks Baltassar for those links. I had seen allocatesmartly’s 10 year Stock Market Return Forecast but hadn’t realized the details of how they created it.

They backtest 20 different published forecasting methods from as far back as they can get data (180’s, 1920’s but most from the early 50’s). A short selection of excerpts from their description of their methods.
"”” We begin with all available indicator data (ex. quarterly CAPE Ratio values). For each of those data points, we determine how the stock market performed in the 10 years following (annualized, dividend-adjusted log return).
We create a simple linear regression for 10-year returns (y) versus the corresponding starting indicator values (x).

We then create a hindsight forecast for each indicator data point based on the slope/intercept from our regression. Forecasts are converted from log to geometric returns.”””

Of the 20 models only 5 end up meeting their positive predictability criteria. From them a weighted prediction is generated.

Impressive amount of work to make a 10-year market prediction but my question is how valuable is a 10 year forecast for an individual investor.

Like many others in the same boat, I have little choice but to try and outperform or at least match the market. Vanguard’s 10-year prediction have Developed markets ex-Us stronger than US but US value still strong.