Extrapolation Can Be Painful: Integral Systems Inc

Author's Avatar
May 05, 2010



Observing trends can be a powerful indicator of future developments in areas such as industry shifts, product trends, commodity prices, etc. However it can be a painful lesson learning the difference between an authentic trend and simply extrapolating recent figures attempting to forecast the future. This is an easy mistake to make for new investors. The past does not equal the future is indeed a powerful maxim. Our human nature convinces us to see the emergence of a pattern of data – even when one doesn’t exist. This falls under the psychological element of illusion of control which deceives us into believing that we can use the past to predict the future. It also exposes another human tendency which is laziness. It’s very easy to take a look at a simple data series and use that as the basis for our investment thesis.


Below, I have highlighted the financials for Integrals Systems (ISYS, Financial) to prove the point.







Book Value/ Share



Debt/ Equity



Return on Equity (%)



Return on Assets (%)



Interest Coverage



09/09



$6.64



0.09



NA



0.6







09/08



$6.38



0.00



16.5



12.3







09/07



$5.68



0.00



12.0



9.4







09/06



$6.46



0.00



8.6



7.4







09/05



$5.78



0.00



5.2



4.2







09/04



$5.30



0.00



6.4



5.1







09/03



$4.92



0.00



5.2



4.1







09/02



$4.41



0.00



3.2



2.7







09/01



$4.31



0.00



5.1



4.4







09/00



$4.33



0.01



4.5



4.1










Financial trends are improving in measures such as ROE and ROA. The natural progression upward can be very tempting. If we had to guess the 2009 result, it wouldn’t be insane to extrapolate improving numbers. Well, for those familiar with Integral Systems, we know that the next chapter didn’t work out well. Granted we were entering an incredibly difficult economic phase. Furthermore, it doesn’t necessarily mean that the improvements weren’t genuine or meaningful. However, it does illustrate that ‘ruler’ type stocks (meaning one could draw the trend line with a ruler) can lead us into a trap. The same principle applies to earnings growth trend. For years names such as General Electric posted numbers that grew on a very even pace over decades. The stocks are rewarded as investors craved the predictability of what was to come next. Yet the divergence from the trend can certainly be a shock to the system – think U.S. housing prices.


It is difficult to avoid the influence of the past and the impact on our thoughts for the future. One possible way to avoid this fate is to ignore forecasting all together. The best way to overcome this bias is to use historical data only as a guide. Try evaluating a business as if today was its first day of operation – it might give you a fresh perspective.