My last blog post focused on some tips on forecasting which were fairly qualitative . I thought that it would be great to follow up by discussing a few technical mistakes done during company valuation. While trying to forecast company performance and valuing companies, we often fall victim of what I would refer to as the ‘routine trap’. That happens when we apply certain methods without understanding the logic behind those assumptions even though situations and circumstances could be very different and require changes in methods. Thus we end up with inconsistent company valuation. Let me go a bit deeper into few common areas of mistakes. Be warned that this will be a fairly long post.
Growing revenues without adequate capex
This is one of the very basic mistakes. By continuously growing revenues we are assuming that the company is increasing capacity. However we fail to model for the capex associated with that capacity increase and hence overestimate company value. This mistake is most dangerous when the capex used in the terminal year is too low and thus an abnormally high ‘free cash flow’ gets compounded using the terminal growth rate.
Typically to solve this problem analysts try to use company guidance for explicitly forecasted (the years where we forecast financial statements) years and use a normalized capex for the terminal year. However, if explicit guidance is not available it is possible to watch capacity utilization and model for capacity additions (costs of incremental capacity can be surveyed from industry participants). A good practice is to keep a watch at RoIC (return over invested capital) and see if that is going up way too much compared to the industry fundamentals. If there is no ‘good’ explanation for ever-growing and high RoIC in forecasted years then the capex estimate is a suspect.
Now, there could be perfectly logical explanations of growing revenues without and major capex (just maintenance capex). A company running at full capacity could be outsourcing some of its production needs to 3rd parties and run a capex light operation. This will lead to higher revenues but usually lower margins. Another example could be a power plant with 15 years life. The bulk of the capex will typically be done at the beginning of the project with only maintenance work being done in next years.
The ability to understand what method to apply in which situation is what will differentiate an experienced an amateur analyst.
Not knowing where revenue growth is coming from
Revenue growth can come from multiple sources. One way to look at revenue is breaking it down in the format prescribed in the book The Granularity of Growth. Their suggestion is to split up revenue growth from volume, price and currency impacts. This allows us to understand what exactly was driving revenues in past years. Failure to understand this could once again lead to incorrect forecasts. Imagine a company in the hotel business or logistics business where it is located outside USA but the revenues are denominated in USD. In an appreciating dollar environment it would see revenues grow significantly just from the currency impact keeping other factors unchanged. If we extrapolate this high growth in a period when USD is getting weaker our forecasts will be grossly of the mark.
The volume growth can be broken down into the natural growth rate of the industry and market share gain and loss. The latter is of prime importance. Imagine an industry which is growing at 10% per year but one specific company is growing at 20%. If we keep on assuming that this company will continue to grow at 20% for a very long period, mathematically we might be assuming that this company is bigger than the industry itself which is illogical.
There are other ways to look at revenue growth also. In the retail industry a key indicator is same store sales which helps to understand how much growth is coming from existing stores and how much from new stores. We can then look deeper into same store sales and see whether foot traffic is increasing or customers purchase basket is the main driving factor.
Not having an interim period for high growth companies
Analysts frequently forecast company financials for 5 years and then use a terminal growth assumption to figure out the value of the company when its a going concern. This is defended on the assumption that our ability to forecast beyond 4-5 years is pretty low.
All this is great when we are working with a mature company. However imagine a dominant consumer food company which is located in a densely populated frontier economy like Bangladesh or Pakistan. It is very likely that such a company will continue to deliver high growth rates for at least another 15-20 years before growth starts coming down. Using a terminal growth rate of 5-6% in the 5th year of forecast will undoubtedly underestimate its growth prospects. What is recommended in Valuation: Measuring and Managing the Value of companies is to have an interim forecast period for such high growth companies where some simple metrics drive numbers between the explicitly modeled phase and the terminal year. The interim period could also have a fadeaway assumption where revenue growth will gradually fade away towards the terminal growth number.
Inconsistency in discount rate and inflation assumption
If confirmation bias is the mother of all biases, then inconsistent discount rate and inflation is the mother of all valuation mistakes. To understand this mistake one needs to understand that the basic premise of valuation is that we are forecasting future expected cash flows with a discount rate to figure out what the future cash flows are worth today. A higher discount rate assumption is either due to a higher real risk free rate, high equity risk premium, high beta or most importantly very high inflation assumption. If we are using a high discount rate DUE to high inflation rates (which gets captured in the risk free rate) then we should be careful about our inflation assumptions in nominal cash flows. Let me explain with an example once again.
The risk free rate in Bangladesh now is around 7%. I however use around 11% (leading to cost of equity of 16%) which in my view in the cross cycle risk free rate. Based on historical numbers I conclude that 11% risk free rate implies inflation of about 8%. Now if in my explicitly forecasted years I am using an inflation rate of just 2-3% then I have an inconsistency problem. The same rule applies in terminal growth assumption. If I am using a terminal growth rate of just 5% I am basically assuming that the company will have no volume growth and will be able to grow prices by less than inflation rate. This will be a wrong assumption.
The key thing to watch out is whether the inflation assumed in the discount rate and the nominal cash flows make sense or not. Certain industries are such that even when inflation is high they are unable to increase prices too much either because of competitive characteristics or regulation. In those, it is actually possible to have high discount rate and low nominal cash flows. We just need to be conscious of the situation and use the right method.
Believing that higher margins lead to high valuations
A specialty toy manufacturer decides to open up a grocery retail business. The margins on the toy business is higher than grocery retailing and thus we forecast that weighted average margins will come down. This much is fine. The mistake occurs when we incorrectly conclude that lower margins will drive down company value. This is because a low margin business can have a very high RoIC and vice versa. The low margin grocery retail business will have much higher capital turnover and whether value of the company will go up or down will depend on factors like execution ability and competitive characteristics of the grocery retail business. Just looking at the declining margins alone will not answer this question.
A good example of this is the telecom sector. The data revenues are growing while voice is saturating or coming down. The increase in data margins means that average margins will be growing. However, data is also more capital-intensive (capex heavy) than voice and thus RoIC is not going to increase just due to the margin increase.
This was a fairly long post but one that I was planning to write for a long time. Feel free to write your criticisms or share your own ideas that can make forecasting and company valuation better.