Limitations of Financial Modeling
Posted on 10. Jul, 2011 by TheFreeInvestor in Investing Philosophy, Valuation
If you have read any of the research reports from wall street firms, you will notice the use of financial models to estimate value of the firm and the target price for the stock. Financial models make us feel good and gives us the comfort in precise numbers and complex formula. The mathematical nature of financial modeling gives us the impression of accurate analysis. Though I use financial models myself, it is important to understand its limitations — which companies it makes sense to model vs others that doesn’t fit nicely into a financial model.
As I mentioned in my post about valuation model, key inputs to the valuation are the projected financials and growth expectations. It is easier to project financials of companies with steadily growing revenue and profits than that of lumpy revenue and profit. Lets look at few typical companies that can be modeled easily. In my portfolio, retail companies like Aeropostale (ARO), Fossil (FOSL) and Guess? (GES) are perfect for financial modeling because they tend to grow in a steady pattern. Similarly, other retail consumer centric companies, e.g., Chipotle (CMG), Starbucks (SBUX) and Fedex (FDX) are good candidates for financial modeling for the same reasons — steady and smooth pattern.
Now, lets look at some of characteristics of the companies or industries that are not ideal candidates for precision financial modeling.
1. Emerging industries and early stage companies
When an industry is in its infancy, it is difficult to estimate financials of a company in that industry. SINA (SINA) is one of my latest investments. It’s a Chinese web portal. But its Weibo service, which is dubbed as the Twitter of China, is the key driver of its value at this point of time. Now, the entire social media industry is still in infancy. There is no clarity about monetization of the Weibo service. So, with so much uncertainty and rapidly changing industry landscape, doing a financial model for next 5 years is pointless. Other companies in my portfolio – EnerNOC (ENOC) in demand response industry and Nuance Communications (NUAN) in speech recognition software market are also examples of similar companies in emerging industries.
2. Industrial companies with lumpy sales cycle
Companies that sell their high priced products and services to other companies tend to have more volatile sales and profits than retail companies. Infinera (INFN) and Dawson Geophysical (DWSN) are such industrials companies in my portfolio (though Infinera probably gets lumped as a technology stock in the market). The product and service they sell have long sales cycle and, therefore, it is tough to project smooth growth rates for financial modeling for next 5 years. Though this can be mitigated by varying the growth rate across the projected years, it still requires a lot of wild guesswork.
3. Industries in flux
This is a difficult one to address. Companies in these type of industries can be modeled using past data, but the reliability of those models are questionable. For example, Netflix (NFLX) in 2006 was an online DVD rental pioneer. Any financial model created in those years will be useless as the company is transforming into a steaming video service today. Though the subscription model is still intact, the pricing and cost structure has changed. In addition, the digital entertainment industry itself is going through a rapid transformation. So, any model created for these companies will require a lot of updates as time passes. Same concept is applicable, may be to a lesser extent, to Amazon (AMZN) or Apple (AAPL). Both companies have transformed significantly in last 5-7 years. So, though it is possible still to build a financial model for these companies, we need to recognize that the model may be less accurate than a model for a retail apparel company due to the rapid innovation in the digital entertainment and eCommerce industries.
Finally, I am not saying that financial modeling is ineffective. I create financial models myself. But, we just need to recognize its limitations and potential effectiveness in different situations. Though a well laid out financial model can be a great marketing tool to demonstrate intellectual depth, it may not produce accurate or reliable estimates in many cases.
(Disclosure: As of the publication of this post, I hold long position in all the companies mentioned above. Please read the full disclaimer on this website.)
Stay Connected
