What Causes Fair Value Estimates to Change?
publication date: Jan 9, 2017
author/source: Brian Nelson, CFA
If you’ve been a member of Valuentum for a while, you’ll notice that when we update a stock report, our estimate of a company’s fair value and the firm’s Valuentum Buying Index ratings can change. This is completely normal and should be expected (over time, companies generate cash). But sometimes the changes can be confusing, particularly if they are material (i.e. 10% or more). In this piece, let’s talk about why changes are standard operating procedure for investment research publishers.
But first, a bird’s eye view of the topic. Our estimate of a company’s fair value is driven by myriad factors. To derive a company’s fair value estimate, we use a detailed cash flow model that considers future expectations of revenue, earnings and cash flows, among other financials (i.e. the balance sheet). You can download the individual-investor-friendly version here (we charge a token price for the template -- hope that is okay). When important drivers change or when new information comes to light, fair value estimates change. Though there are perhaps an infinite number of reasons why a fair value estimate can change, there are two primary reasons that account for the vast majority of revisions.
Rolling the Model Forward
The biggest fair value revision occurs when “we roll a company’s model” forward one year. This is analyst speak for when Year 1 of the model changes from, say, 2015 to 2016, or in the case of what will happen next year, from 2017 to 2018. The timing of this revision occurs after a firm issues its fiscal annual report (form 10-K or form 20-F). For most companies, this occurs late in the first quarter, and our updating proceeds through the second calendar quarter. Once we receive the audited new information for the last fiscal year (which is released in the form 10-K or form 20-F), the new data is entered into the model.
This is why you may see last year's data still listed as Year 1 through most of the first quarter of any year. We're still waiting for the clean, audited numbers.
Generally speaking, a company’s fair value estimate will increase by its discount rate less the dividend yield each year, all else equal. However, this increase almost never happens in practice, as the trajectory of the firm’s future free cash flow stream and its capital structure are refined with the new information in the 10-K or 20-F. For example, if a company has engaged in value-destructive activities during the previous year (e.g. it has overpaid for acquisitions and bought back its own stock at egregious prices), this will show up in the new fair value estimate. On the other hand, if a firm is a wise capital allocator, the firm’s balance sheet and future cash flow trajectory will have been enhanced from the previous year, and this would cause an upward revision in the fair value estimate (sometimes by 10% or more), all else equal.
When “rolling the model forward,” there are a near-infinite number of drivers that could influence the fair value estimate, though we point to changes in the balance sheet (specifically the net cash/debt position) and changes in the future free cash flow stream (revenue, EBI, capital spending, working capital and other components) as being the two biggest factors. Most of the drivers behind a change in a fair value estimate, resulting from when “we roll the model,” will be operational (e.g. updating the cash flow trajectory and accounting for cash generated during the previous year as reflected in the updated balance sheet and/or lower share count). We tend not to adjust a firm’s cost of equity, nor do we adjust the risk-free rate frequently, though this may happen in some cases.
Note: Our model does not account for cash generated in interim model updates during the year. For simplicity, we account for cash generated during the previous year once, at the time we “roll the model”-- not continuously.
Significant Changes in Expectations or Transformative Acquisitions
Valuentum’s fair value estimates, and by extension, the Valuentum Buying Index rating are forward looking. That means when expectations of a company’s future free cash flow stream are revised as a result of forward guidance revisions (or incremental insight from our analyst team), or when a company pursues a transformative acquisition that will materially change its capital structure in the future, the fair value estimate changes accordingly.
The variables that cause the biggest changes in the fair value estimate on an operating level are our forecasts of a company’s mid-cycle operating margin (Year 5) expectations, mid-cycle (Year 5) revenue growth rate, and capital spending over the 5-year discrete forecast period (or phase I of the model). If firm XYZ, for example, comes out with substantially lower revenue and earnings guidance for Year 1 than what we and the Street had been modeling, there may be a downward revision in the company’s fair value estimate, all else equal.
Though Year 1 (or even Year 2) in the model does not impact the fair value estimate materially in most cases, the information behind the revised guidance could influence the intermediate-term and even the long-term forecasts of the model (think ripple effect), and this would cause an even larger fair value estimate revision (in some cases). Whenever the trajectory of the future free cash flow stream changes, the fair value estimate, which is based on the future free cash flows, changes. This is a fact of value calculation.
How to Think About All of This
First of all, we are not a quant firm. What you see in our reports is in-depth fundamental financial analysis. Analysts from investment bankers to sell-side brokerage firms use a discounted cash-flow model. Their results may be presented in the form of a multiple, but every merger or business transaction is backed by a pro-forma discounted cash flow model that justifies the transaction multiple. Conversely, the multiple implicitly considers all of the information in a discounted cash-flow model (you can read more about this relationship here). Even a PE multiple applied to this year’s earnings has an implied growth rate for earnings 10, 15, 30 years…into the future. Thinking about the long-term in valuation is inescapable, and the discounted cash-flow model is the best way to think about the long-term.
Second, we don’t just hit a button and the published fair value estimate pops out of the model. Though we understand that being roughly right is better than being precisely wrong (think in the context of Warren Buffett's margin of safety), we spend a tremendous amount of time modeling companies. We don’t play the near-term earnings estimate game, however. For one, we’re fairly certain we can’t do better than the 40 or so sell-side analysts that cover the average S&P 500 company these days, so our near-term forecasts don’t differ much from management’s guidance. On average, our revenue and earnings estimates for the next two years won’t be but a couple percentage points away from consensus estimates. You can read more about this topic here.
So If We're Not Playing The Earnings Game, What's Our Advantage?
Our advantage rests in our modeling expertise and synthesizing the framework down to a few key drivers within the detailed free cash flow model that we think the Street often gets wrong (namely mid-cycle operating margins and mid-cycle revenue growth). Whereas the Street spends most of its time playing the quarterly earnings game, we’re instead focused on nailing our mid-cycle forecasts and getting the company’s fair value estimate correct. Most of a company’s intrinsic value is based on the structure of its balance sheet and its future expected normalized free cash flow stream (earnings before interest less net new investment), not on next quarter’s earnings.
We understand that there may be a 10%-20% revision or more in a company’s fair value estimate at times, but this is normal in the world of value estimation (expectations of the future are always changing and sometimes materially). You can always view the ‘Valuation’ page in a company’s report to see exactly what forecasts have changed from the previous report (page 5 gives you a very nice summary of the three stages of the model). In almost all cases, the updated fair value estimate will fall within the previous report’s fair value range.
We embrace the concept of Benjamin Graham's margin of safety, and we consider firms to be undervalued or overvalued when its share price is trading below or above the fair value range, respectively (though, at times, we may add firms trading at less of a discount that we may desire -- good companies sometimes never trade at bargain-basement prices). In the 16-page valuation reports, the percentage difference from the company’s share price to the low end of the fair value range is considered the ‘percentage undervalued,’ and the percentage difference from the company’s share price to the high end of the fair value range is considered the ‘percentage overvalued.’
The very best way to get a feel for all the moving parts of valuation is to take our valuation model for a test spin. You don’t have to become an expert on this process to do well -- just thinking about the key variables behind a discounted cash-flow model will make you a better investor, in our view. We hope you’ve found this article informative, and if you have any questions about any aspect of our research, please don’t hesitate to reach out to our team.
How frequently are the (stock and dividend) reports updated and what triggers an update?
The 16-page stock and dividend reports are updated at least every 3-4 months or when material results alter our estimate of a company's fair value. This update cycle is typical for investment research firms. Fair value estimates do not change much over short-term periods of time, especially if our estimate of a company's intrinsic value is spot on. Although the date, data, or text on our reports may not change daily for each company, one can assume that if the reports are live on the site, the conclusions of the company in the report are still representative of our view on the stock or its dividend.
Importantly, by using our 'Symbol' search box in our header, one can gain access to some of the most advanced charting features to augment our Valuentum Buying Index's technical/momentum assessment. The charts are equipped with real-time data. The 'Symbol' search box in our website header is also the best way to find our recent articles and analysis (as well as the 16-page equity and dividend reports) on companies of interest to you.
Pasted below is what the search box retrieves on information related to Intel (INTC), for example:
A version of this article appeared on the website June 27, 2014.