This recession (or worse) is wreaking havoc on most companies. Their financial statements will reflect it as long as it lasts. For sure, 2008 and 2009 results will suffer, and maybe those for 2010 and beyond. Since we typically go back five years when analyzing historical financial results, we will be considering 2008 at least through 2013, 2009 through 2014, and so forth.
When we analyze historical financials, we normalize them to try to reveal the true earning (more accurately, the cash generating) potential of a business. This assumes that the future will be like the (normalized) past…obviously a dangerous proposition given what is going on in the world today. Change has indeed come, to quote the President, and this may invalidate historical comparisons regardless of our ability to normalize. If you don’t believe me, would you consider General Motors’ historical financial results for the past five years, normalized up the wazoo, to be valid for any projective purpose?
(Or, as my son in-law, an inveterate gambler, likes to say, “Dice have no memory”.)
Closely tied to the concept of normalization is the concept of weighting historical results. One common scheme assigns last year a weight of 5, and previous years 4,3,2, and 1. Under this scheme, last year weighs in at 33% (5 of 15), the middle year weighs in at 20%, and so on. Alternatively, you could give a truly atypical year a zero weight (rather than try to normalize it, which might be impossible). The problem with any weighting is that it is totally subjective, and the weights have a huge influence on the result. I do not recommend the use of weighted historical averages.
Back to normalization, there are many reasons why actual results might need to be adjusted:
1. To go from actual results (benefits available to minority owners) to control results. Examples occur when we add back excess compensation and adjust for non-arms length transactions such as below-market leases to affiliated parties.
2. To remove the effects of one-time or non-recurring items. Examples are gains or losses on asset sales or casualty losses. Sometimes these are not straightforward; maybe the effects extend over more than one accounting year. Also, every business has some degree of non-recurring and one-time expenses every year. To adjust to a “perfect” year is probably going too far.
3. Assets (and liabilities) recorded at historical cost may need to be marked to market. We ask fixed asset appraisers to help with real estate and equipment. Inventory may need to be adjusted (LIFO to FIFO, for example). Investments may need to be revalued. And fair value accounting requires that intangible assets and goodwill be identified and valued.
4. Accounting may need to be converted from cash to accrual.
5. If the integrity of the financials is questionable, forensic adjustments may be needed.
6. There may be new tax laws.
7. In acquisitions, we may want to assess the impact of synergies.
I may have missed other valid reasons for normalization – please let me know what I left out. But there is one thing we should NOT normalize for: the performance of existing ownership/management (if there is to be no change of control). Particularly in the case of underperforming businesses (compared to industry norms), it is not correct to blithely increase (for example) historical gross profit margin for any reason other than those cited above. If owner/managers have figured out some way to increase margin, that should be reflected (if you agree with them and with potential adjustments for risk to the discount rate) in forecast results.
There is one exception to this suggestion: in divorce cases, businesses tend to begin to under-perform when the marital parties get into trouble…these results need to be taken with a grain of salt.