If a valuation multiple, such as EV/EBITDA, is used to calculate a DCF terminal value, the multiple should reflect expected business dynamics at the end of the explicit forecast period and not at the valuation date. This is best achieved by basing the exit multiple on forward-priced multiples for the selected group of comparable companies.
We explain and illustrate with an interactive model the use of forward-priced multiples in DCF. We also discuss the choice of multiple (including why EV/EBITDA may not be the best) and whether to apply the exit multiple to reported or adjusted profit.
DCF valuations typically involve an explicit forecast of cash flows, with value beyond that period summarised through a terminal value calculation. A terminal value can either be based on a continued, but simplified, cash flow calculation, often using a perpetuity growth rate, or it can reflect an assumed multiple applied to the explicit forecast profit at the beginning of the terminal period.
In our article DCF terminal values: Returns, growth and intangible assets we discussed some aspects of cash flow based terminal values, in particular the relationship between reinvestment rates, returns and growth. In this article we focus on the application of exit multiples based on the analysis of comparable companies. We consider three issues:
- Why it is important that exit multiples are derived from forward-priced multiples of comparable companies
- Which multiple is best
- Whether exit multiples should be based on adjusted (non-IFRS or non-GAAP) performance measures.
Comparable company multiplies should be forward priced
Where an exit multiple is based on comparable companies the characteristics of those peer group companies should, as closely as possible, match the characteristics of the company to be valued. This means selecting companies that have the same business activities and are in a similar stage of development, and that consequently should have similar future growth and return expectations. However, by itself, this is not enough to obtain a suitable exit multiple.
Terminal multiples should reflect expected business dynamics at the start of the terminal period
The objective of the comparable company analysis is to identify, for the company you are valuing, what multiple it is likely to trade on (based on expectations at the valuation date) at the end of the explicit forecast period. This means that the business dynamics or value drivers reflected in the multiples for the peer group should be consistent with those of the valued company at the start of the terminal period, not at the valuation date. It is for this reason that we prefer to use ‘forward-priced’ multiples to derive a comparable company based terminal multiple.
A forward-priced multiple is essentially the terminal exit multiple implied by the current observed market enterprise value (the EV based on current market capitalisation) after considering the other components of an enterprise free cash flow DCF valuation. Fortunately, a full DCF model is not required for each comparable company.
A forward enterprise value is the current market EV, less the present value of enterprise free cash flows forecast for the intervening period (the explicit forecast period when used for a terminal value) with the result compounded at the cost of capital for the required number of years. An approximate and simpler method, and the approach we apply in the interactive model below, is to apply an assumed free cash flow yield instead of an explicit cash flow forecast. The forward EV is the current EV compounded by the difference between the cost of capital and the free cash flow yield.
The forward priced multiple is the calculated forward EV divided by a forecast metric such as EBITDA for a related forward period. For example, a ‘year 5 first year prospective’ EV/EBITDA multiple would be the forward EV for year 5 divided by forecast EBITDA for year 6.
Forward priced multiples – An example
Assume that the current market enterprise value for a company is 1,000 and the forecast first year prospective EBITDA is 100, therefore the current priced ‘1E’ EV/EBITDA multiple is 10x. In the next three years the enterprise free cash flow is projected to be 50, 70 and 90 respectively, and that the weighted average cost of capital is 9%. The forecast EBITDA for year 4 is 180.
To calculate a year 3 forward priced 1E prospective EV/EBITDA multiple we must compare the year 4 forecast EBITDA with the year 3 forward enterprise value where the forward EV is the implied terminal value of a 3 year DCF valuation that reconciles to the current EV.
An approximation of this forward multiple can be calculated by applying a free cash flow yield in place of the explicit forecast of cash flows.
Forward EV = Current EV * (1 + WACC)n / (1 + FCF yield)n
The approximate approach uses the current free cash flow yield as a proxy for the present value of expected cash flows over the next three years. Because FCF is rising, the yield approach will understate the cash flow adjustment and hence overstate the forward EV and resulting multiple. A better approach would be to estimate an average FCF yield for the forward period (3 years in this case). Generally, the FCF yield approach should produce forward priced multiples that do not materially differ from those using an explicit cash flow forecast, especially if the forward look period is relatively short.
The model below is an extract from a larger downloadable spreadsheet in which we provide five alternative approaches to the calculation of a DCF terminal value. The calculation features a set of comparable companies with data for their current EV and forecast year 1 profit. We have used NOPAT although other EV related metrics could also be used.
Interactive model: DCF terminal value using forward priced multiples
Note: In addition to applying the simplified forward priced multiple approach we illustrated earlier we also simplify the forward profit metric calculation by applying a single compound growth rate to the year 1 profit.
Please enter your email address to receive an excel version of this model
Several of our comparable companies have high forecast growth over the period to year 5, the terminal year of the explicit forecast in our DCF model (based on the default data when you first load the page). This means that the year 1 prospective multiple will be high and not suitable as an exit multiple for the company being valued, where we have an assumption of long-term growth of just 2%. For example, comparable company 1 has a year 1 prospective EV/NOPAT multiple of 20x. However, when this is rolled forward using the forward pricing approach, the 1-year prospective year 5 multiple (year 6 forecast profit compared with a forward priced year 5 EV) is only 9.8x. This ‘ex-growth’ exit multiple is a much more appropriate exit multiple than the current prospective multiple.
The result of forward pricing is a median multiple of 11.0x for the 7 comparable companies. This compares with a median current multiple of 19.0x, the application of which would have led to a significant over valuation.
Forward priced comparable company multiples also allow for differences in cash flow generation and cost of capital during the explicit forecast period, which further enhances their relevance for calculating terminal values. For more about forward priced multiples, and for a model illustrating alternative approaches to their calculation, see our article Why you should ‘forward-price’ valuation multiples.
A lack of forward pricing in practice – Tiffany
Fairness opinions provided at the time of business acquisitions generally include DCF valuations and often these feature terminal values based on observed comparable company multiples. The extracts below are from the offer document for the purchase of Tiffany by LVMH in 2020. This includes a fairness opinion provided by advisory firm Centerview who base their opinion on a comparable company multiple valuation, using both EV/EBITDA and P/E ratios, and a DCF valuation that includes a 7-year explicit forecast from 2020 to 2026 followed by a terminal value based on an exit EV/EBITDA multiple applied to the 2026 forecast.
The 2020 EV/EBITDA of the identified comparable companies range from 6.6x to 21.7x. For the multiple based valuation, Centerview applies a multiple range of 11x to 17x to 2020 EBITDA. However, for the 2026 exit multiple in the DCF valuation the range is 13x to 18x. There are no explanations of the basis for determining the exit multiple range and whether it is based on the same 2020 multiples or a separate calculation of 2026 forward priced comparable company data.
In our view, using the same current-priced comparables data for a multiple-based valuation today and for a DCF valuation that includes a terminal value based on a year 6 exit multiple, is only appropriate if the business dynamics and value drivers are forecast to be unchanged over that 6-year period. In other words, the expected value drivers at the end of 2019 as reflected in 2020 EBITDA multiple should be broadly similar to the expected value drivers at the end of 2025 to justify a broadly similar 2026 exit multiple. A much better approach is to calculate forward-priced multiples for the comparable peer group to use in the DCF terminal value.
Extracts from Centerview fairness opinion on the valuation of Tiffany
Comparable company data
Multiples based valuation
DCF based valuation
Source: LVMH offer document for Tiffany
Of course, we are not saying that there is necessarily anything wrong with the fairness opinion. Centerview may well have allowed for the expected change in business dynamics over the next 5 years in other ways. Indeed, they say that the exit multiple was selected “utilizing its professional judgement and expertise”. However, in our view, the analysis and the rationale for selecting the stated exit multiple range would have been clearer if the forward price multiples of the comparable companies had been presented.
Which multiple?
There is no right answer to the question of which multiple is the best basis for a terminal value in an enterprise DCF valuation. Any profit, cash flow, asset or indeed other activity metric could conceivably be used. Our preference would be to always use profit rather than cash flow and a measure that is comprehensive and therefore differentiates based on all factors that affect overall profitability. For this reason, we would generally prefer EV/NOPAT, except modified to exclude the amortisation of ‘replacement expensed’ intangibles arising from a business combination (see below).
EV/EBITDA may not be the best multiple to use for terminal values
In practice it seems that many investors prefer to use EV/EBITDA. Indeed, this appears to be the norm in fairness opinions (including in the case of Tiffany above). In our view, this multiple has serious shortcomings due to the exclusion of depreciation and taxation. Effective tax rates and particularly capital intensity (the level of maintenance capital expenditure relative to other operating expenses) can vary considerably, even amongst companies in the same peer group. Both these factors affect value and hence the deserved EV/EBITDA multiple. Taxation directly affects value because it is a cash outflow – either immediate or deferred. Depreciation affects value indirectly because it can be seen as a proxy for maintenance capital expenditure.
The lack of a direct link between depreciation and value leads some to prefer to use EV/EBITDA and allow for other factors, such as maintenance capital expenditure differences, through the judgements made in selecting the exit multiple. However, in our view, it is best to build the differences in capital intensity in the multiple used. This means using EV/NOPAT or, even better, use an adjusted form of NOPAT in which the depreciation is replaced by estimated maintenance capital expenditure.
Adjusted or unadjusted profit?
Many investors calculate valuation multiples based on adjusted (non-IFRS or non-GAAP) profit metrics. The use of adjusted metrics reflects the fact that guidance provided by management is generally based on the adjusted management view of historical performance. While adjusted metrics may give a better view on historical performance due to the exclusion of certain non-recurring items, there is a danger that, in DCF analysis, forecast profit and cash flow, and therefore terminal values, may be overstated. Adjusted metrics are often incomplete and exclude items that, while difficult to forecast, would nevertheless have a non-zero expected value that therefore affect valuation.
Expenses excluded from management guidance may still be relevant for DCF terminal values
Restructuring costs, asset impairments and other so-called non-recurring items may be excluded from adjusted profit due to their volatile nature. We think that this is often not appropriate for profit and cash flow used in DCF. In such cases forecasts will be difficult but, considering, for example, the average of these expenses in prior years, should facilitate a better forecast than simply using zero. Obviously, impairments are a non-cash item, so including these in NOPAT and then adding it back again when computing free cash flow might seem pointless. However, impairments of assets with a limited useful life are essentially a catch-up of prior under-depreciation and, to the extent that terminal values are based on profit multiples or reflect a return forecast, we think that including these impairments as part of the profit forecast is important.
It is also common to see stock-based compensation excluded from adjusted profit. We strongly disagree with this adjustment for performance measurement and even more so for DCF valuation. The argument that SBC is non-cash, and therefore cannot affect DCF values, is false. SBC is an example of an ‘effective’ operating cash outflow. It is essential to include forecast SBC in free cash flow AND deduct the value of outstanding stock options in the bridge to equity value to fully reflect this in a target stock price. For more about effective cash flows and the analytical implications of stock-based compensation see our articles When cash flows should include ‘non-cash flows’ and Dot-com bubble accounting still going strong.
Although we think that IFRS or GAAP forecasts are generally more suitable than management adjusted metrics as a basis for valuation, there is one adjustment we think you generally should make – the add back of the amortisation related to so-called ‘replacement expensed’ intangibles arising from a business combination. The add back depends on whether you have made comprehensive adjustments to capitalise all intangibles, as we highlighted in our article Missing intangible assets distorts return on capital. If you have done so, leave in the acquisition related amortisation; if you have not, make the add back. As with impairments, amortisation is non-cash, so it may seem an irrelevant adjustment in DCF, but we think this adjustment is essential.
Insights for investors
- Exit multiples used to derive DCF terminal values should reflect the expected characteristics of the company being valued at the point of that terminal value, not at the valuation date.
- Use forward priced multiples for comparable companies to ensure the most relevant terminal value multiple.
- Don’t use EV/EBITDA as an exit multiple just because everybody else seems to. Also consider EV/NOPAT, particularly with the added maintenance capital expenditure adjustment.
- If you choose to use EV/EBITDA as a terminal multiple be aware that taxation and capital intensity are key drivers and differences can affect the relevance of comparable company-based multiples.
- Profit forecasts used in DCF terminal values should be based on GAAP not non-GAAP performance, except that you should exclude amortisation of ‘replacement-expensed’ intangibles, unless making comprehensive adjustments to capitalise all estimated intangible investment.
Don’t miss new articles published by The Footnotes Analyst – subscribe here:
No advertising, no spam, just new articles. Unsubscribe anytime.
Send us a question or comment about this article:
For more articles select subject tags:
Or see suggested related articles below: