Financial models and their nuances (Part 2)

Continuing from our previous post on Financial models…

In this post we talk about the other two components of Financial models. The reason we have separated these posts is to differentiate the “Science” from the “Art”. We believe that Historical data and the Business Model represent the science of investing. They are quantifiable and objective in nature. Whereas the Assumptions for future drivers and valuations are more of Art. Future assumptions are always subjective in nature and can be debatable – perception for the future outlook of businesses can vary between investors, management and other stakeholders. Similarly valuation is a fluidic topic and is a subject of great debate and argument between investors. As such valuing stock prices depends greatly on the inputs that one provides for these two aspects of financial modelling.

Historical data and the Business Model represent the science of investing -they are quantifiable and objective; Whereas the Assumptions for drivers and valuations are more of Art – as they subjective in nature

We had talked about key drivers in building the business model of the company. These drivers interact with each other and historical values to help us estimate the numbers going forward. For a Financial model this where the 3rd key aspect arises – what values to enter for the drivers? Once you enter the values of these drivers, the model will throw up the expected financial performance of the company going forward. So if your expectations for the drivers are incorrect, the expected performance will also be incorrect and in most likelihood so will be your estimate of the fair value of the stock, leading to an erroneous investment decision.

The quality of analysis and accuracy with which analysts arrive at the forecasts for these drivers is what sets apart a good analyst/investor from a one not so good. In essence how do you say that a company is likely to deliver a 5% growth or 15% growth in the next few years. This is by itself an expectation of the future and there is no way to know how accurate it is or will be. In order to work with this limitation, the market always works with expectations. So the market price by and large reflects the consensus expectations of the broader market at large and analysts and investors work to uncover any likely deviations in actual performance from the consensus.

This is where we believe our models will add value to the market. Let us take a few examples. A cement dealer will have a good handle on on-ground demand and pricing and more likely to have better estimates for demand growth and pricing for cement companies. He can use our models to see the possible financial performance of these companies and then compare them with what the market is already pricing in. Similarly there are consultants who have a good top-down view of the industry and are likely to have better estimates than most others, they can use these models to value stocks better. A fund manager/stock analyst can take individual inputs on specific drivers from each of these sources to create their own expectations of stock prices and manage their funds appropriately.

So as we have said earlier as well – DistrictD’s job is solving the science of investing, enabling better collaboration and making markets more efficient. We leave the decision making and the Art of investing to the investors.  While the quality of a financial model effectively depends on how accurate the assumptions for the drivers, one needs to keep in mind the following issues as while factoring in their expectations for the drivers

  1. The overall picture – To explain I will start with an example. An investor A, is positive on a Company X as he feels the volume growth will be a stunning 50%. While he is correct on it on a standalone basis, he misses the fact that the company did heavy price discounting to achieve the same. In fact X had dropped their pricing by 33% to gain market share. As a result even though A was correct to an extent, X’s revenues remained flat as against A’s expectations of 50% growth.
    As such volume growth and pricing might be independent of each other only to an extent, beyond that they are mutually dependent (think price elasticity) and as an analyst one would need to factor that dependence between the drivers appropriately. This is valid for every other driver as well – such as marketing spends, cost of delivering services etc.
    As such an analyst needs to look at the all the drivers in totality to be able to make a properly tuned model and get to realistic expectations of the future performance of the company.
  2. Evolving business model – Company’s business models keep evolving over time. Our models are based on the quarterly and annual disclosures that a Company makes. Because there is a lagging effect, there maybe some deviations in the business model, and as such with every quarter or major change in the business, the business model has to be adjusted/tweaked. While we do try to ensure that the business model that is captured in our models is a best effort modelling of the company, there maybe some deviations due to the ever evolving landscape. In such a case it becomes important for the analyst to factor in such evolution in the input parameters appropriately to get to the right financial estimations.
  3. Sensitivity – One of the key use cases for financial models is to see the sensitivity of various drivers on the financial performance. However as indicated the drivers themselves might not be mutually independent. This means that changing an assumption by a slight amount will necessitate a corresponding change in some other factor as well. So one needs to be careful while performing such analysis. A large part of this phenomenon is because of management discretion, whereby they look at the overall picture while running the business and tend to move several drivers at the same time to achieve desired goals.
    In our models, as far as possible we try to incorporate drivers in such a way that they hold their mutual independence w.r.t. all other drivers (atleast for minute changes). This allows users to atleast perform sensitivity to incremental changes. In order to model large changes better, it is always advisable to look at the overall picture and see how some of the other drivers are also likely to change in a competitive market.

Once the driver assumptions have been set, the model gives us the expected financial performance of the company. Now comes the last component of the model – its valuations. How do we value the company, both in terms of process and the value attributed to it. Valuation is a huge topic on which people have filled mountains of paper. For the purposes of this post, suffices to say that our models have inbuilt valuation tools where we allow users to choose their methodology as they see fit and a number of supporting features that allows one to do valuation analysis like a professional investor would do.

As such once the four components of the model are inplace the model is completed, with the end result being a fair value for the company’s stock. At our platform, one can also note down their investment thesis for the model and save/share it as they deem it fit. We advise creating an investment thesis along with the model as it helps articulate the thought process behind the driver assumptions and the valuations based on the prevailing market conditions. This often helps one to compare the thought process in the past with the actual conditions in future – allowing the investor to see where he went right/wrong on the drivers.

Leave a Reply