January 21st, 2008, 11:58 pm
One option is to look at relative valuation versus the whole population. If you have any data about line of business at all, you can throw it into your analysis and break down the list by sector. It's fun to use regression studies to see what actually explains the variability in the observed multiples (country, sector, historical/projected growth, size, ...).You can also coble together some sort of DCF based on industry average growth rates, or by projecting past growth rate forward over a finite horizon before taking a terminal value.You could also look at an excess earnings approach, which primarily depends on book value and return on assets.In the sloppiest of all worlds, a company is worth:V = E1*(1-k)/(r - ROE*k)V == valueE1 == Earnings in the next period, which is like trailing earnings projected ahead one year. The yearly growth rate of earnings is ROE*k in this model.k == Earnings Retained In Each Period (meaning the % of earnings reinvested in the business and therefore adding to next period's Book Value)ROE == Return on Equityr == discount rateFrom the description of your database, I'm guessing that you will have a lot of special cases to deal with. Good luck.You might want to look in a couple of Damodaran's books, he's a prolific NYU professor who works in this area.