In September 2006, the IMF in its Global Economic Output report projected that the US will be “the engine of global economic growth despite some uncertainty in the housing sector”.  In its July 2008, less than 2 years thereafter, the IMF revised downward (again) its projections for US economic growth to 1.3% (from 3.2% and 2.2% in the previous two semi-annual reports).  The world most sophisticated economic forecaster failed to project the size and the impact of one of the most serious financial crisis in our modern history on the most monitored economy, although that event was on the radar screen for two years.

Last week I spent a few hours with a colleague of mine trying to “fit” the actual financial results of a company we had invested in into the model we had built for that company prior to that investment.  The bottom-line projections were close to the actual results, yet we couldn’t put the right parameters in the model that will get us even close to reality.

It was an intellectual exercise, but at the end I had to pause and think: if a model cannot predict the past, how on earth we were confident that it could predict the future?

Make no mistake, this is a universal problem and is not a result of any individual incompetency. 

Thousands of financial analysts pour their energy to predict the “fair” value of stocks and other financial assets.  They calculate with pinpoint accuracy the price of the stock after going through the mundane task of calculating things like future revenues, overheads, profits, discount rates, betas, comps, etc.  Yet, with all the sophisticated modeling, I have rarely seen a numerical analysis explaining or predicting the “actual” market price of a stock. Furthermore, research has shown that actual prices do not necessarily gravitate towards “fair” prices as the theory behind all this analysis suggests.  In other words, the thousand of “fair” value models published by investment banks invariably fail to predict the present (or the past) and are unlikely to predict the future.

Yet we rely on projections to give us confidence in our decision. No investment is done in the modern financial world without a model. No model is without future projections. We project profits in 2015 to be $56,405,383.34, yet the only fact we can be sure of is that they are not. We base our decisions on such numbers, and we focus our attention on things like the IRR as the analytical summary of thousands of assumptions and projections.  We may reject an investment because the model calculated an IRR of 29%, while we all know that the same model with a little bit of subjective tweaking and twisting may give a magical boost to the IRR to, say, 39%.

Prior to inventing Lotus 123 and its successor MS Excel (and before I was born), investors relied on qualitative assessment and simple calculation to make their investment decisions. Did they take wrong decisions? Does a 1000 line model give an edge over such “unsophisticated” investors?

Naseem Taleb inspiring best seller “The Black Swan” argued skillfully that we simply cannot predict the future because “significant yet improbable events” (ie he tagged them black swan events) are the ones that will shape the future of individuals, countries, companies, and economies.  We simple cannot predict or time or imagine such events, and we definitely cannot project their impact (hence the IMF miscalculation)!

Projections are useful to the extent they are used as an analytical tool for possible future options. They may bring illusionary confidence, but reality will surely be different, especially in private equity investments that span several years. Embracing this fact as we price, negotiate, structure, and manage any investment is the key to sustained successful investment.