In an uncertain world, the question is always the same - whose opinions do you trust? Where do you put your forecasting faith? Is it a person? a model? specific data?
Clearly, forecasting is always hard with the potential for error increasing with uncertainty. Hence, there must be some place or base that can be used for starting the forecast or forming a null of what may happen in the future. For those who focus on market efficiency, the base is always the current price. Your best guess for tomorrow is the price today.
I try and make this simple with a core rule for any prediction discuss. Let prices (trends) speak for themselves. The direction in prices is the null or base for any for forecast, not the current level. If prices are moving higher (lower), then our base case is that prices will continue to move higher (lower). Of course, this single direction cannot continue forever. That is not the point. The point is that any forecast other than the trend needs to have a strong justification.
Trends can take prices away from fair valuation, but the first view is that trends are pushing prices to a new valuation. If they continue to trend, then the market has not taken it to this new valuation. Prices move toward the new fair value not back to an old fair value as measured by a past price level. You may forecast that the speed of adjustment is faster or slower, but the first pass is that prices are telling you something about the aggregate behavior of investors or the weighted opinion of speculators. There will be deviations from a base for fair value, but a sustained price move is forecast for a new valuation. The fact that a trend-follower uses a longer-back test is just a process for eliminating short-term noise.
Let the data speak first, then adjust forecasts only if you believe you have an edge with incorporating new information.
No comments:
Post a Comment