Seven Deadly Sins II: Multiple Shenanigans

A month or so ago, I wrote an article on the seven deadly sins of DCF valuations. Since that time, a number of people have highlighted to me that these valuation deadly sins don't only occur in DCFs.

They also arise in comparable multiple valuations (CMV). Accordingly, I've come up with another list of seven, where these sins are committed in CMV.

But multiples aren’t all bad

Before we get into the sins, CMVs have a pretty dodgy reputation – one that is mostly undeserved.?Whilst the concept applies simple mathematics developed to handle valuations in the days before spreadsheets made DCF valuation and simulations really easy, CMVs still serve well in many applications.

The simplicity reduces the error rate (only two things can go wrong) and magnitude (errors are only multiplied once not the many times of a DCF model).?They also are useful in dispute contexts as they reduce the fronts of argument to just two, rather than the matrix of arguments that a DCF represent.?They are also theoretically consistent with a DCF, albeit one with only a terminal year.

However, they have limitations.?Because CMVs are really only a relative valuation methodology, they are most useful in estimating price (after all, that is what most professional valuers are tasked with).?CMV though are not great at measuring the most important variable for all deal doers, value; i.e. what you get from the deal.?The DCF approach allows modelling of exactly how you are going to derive value from the asset, and, accordingly, with sensible assumptions is vastly superior to multiple valuations for assessing value, as opposed to price.

As long as you keep their limitations in mind, multiples can be very useful.?Provided that is you don’t let the following deadly sins creep in.

Sin #1: Hardly normal

Normalisations is a process which eliminates unusual or one-off items in an historical or forecast earnings figure in order for a multiple to be applied. This is a process which is a big part of due diligence efforts all around the globe.

However, in valuation exercises for non transaction purposes, sometimes the process of normalization is cursory, at best. Accordingly, things that inflate or deflate earnings in one particular period can be multiplied by the earnings multiple that is used. And this can compound a one-off issue or inflate the effect of a one-off windfall.

Sin # 2: Bad weddings

Bad weddings arise when the wrong multiple is applied to the wrong earnings.

This can arise in several ways. The most obvious is applying an historical multiple to forecast earnings. Usually, forecasts incorporate an element of growth, which means that forecast multiples when used are typically lower than historic ones. So, applying a forecast multiple to historic earnings would typically overinflate value.

In a similar vein, applying EBIT multiples to EBITDA massively inflates the value of companies by avoiding the impact of capital spend in the earning stream.

Sometimes too multiples can reflect different time periods, especially around the reporting dates for the companies.?For example, just before?reporting date an FY+1 multiple may actually be effectively a historical multiple, failing to address this issue and “calendarise” the multiples can lead you astray.

These bad weddings pervade many valuations, particularly by less experienced practitioners.

Sin #3: Control premiums and the flaw of averages

Many valuations will record a control premium, typically somewhere in the order of 20 to 25% control premium, quoting the average of multiple studies, which conclude that the control premium observed on average over time is within these sorts of ranges.

However, views expressed in this way tend to overlook the fact that whilst the average may be this, and consistently so, there is considerable variation around the average.

Control premiums can be zero, in some circumstances, and in others, the control premiums can be up to a hundred percent. Clearly, using the average in this context would only be appropriate if you're looking at an average transaction.

More sophisticated valuation analysis takes into account the circumstances of the change in control, and allows for a differential control premium, depending on circumstances. One of my colleagues on the valuation committee of the Institute of Chartered Accountants recently wrote an article exploring just these sorts of issues.

Moreover, at the current time transactions are often executed using a dual track approach, and often result in an exit being achieved by an IPO, rather than a control transaction.?In these situations it is difficult to argue that a control premium would apply at all.

Sin #4: Multiples that just ain’t the same

When applying a multiple in an industry, it's common valuation practice to take the average of the list of identified guideline companies. Like for control premiums, averages reveal some things, but conceal many others.

Accordingly, it's best practice to look at what drives the differences between the various company multiples. For example, when looking at sales multiples, often the most significant difference is the profitability that the company achieves from their sales.

In a similar vein, EBITDA multiples are often highly affected by both the capital intensity of the business and the future growth possibilities of that business.

Moreover, PE multiples are massively affected by the level of leverage and depreciation policies of the subject companies, and these need to be allowed for. If not, the application of an average multiple could be exceedingly misleading.

Another trick for multiples is finite life businesses, like mines or government concessions.?The life of the asset will drive a major difference between businesses in these situations too.

And finally when it comes to CMV; size really does matter.?Applying a multiple drawn from, say, the ASX100 and applying it without adjustment to a business with a turnover of a few million dollars is likely to significantly overstate value.

Sin #5: No cross checking

As in DCF approaches, cross-checking is often a way to minimize the errors in terms of valuation.

Of course, if, when doing a CMV, you have the luxury of being able to do a DCF valuation using forecasts derived by the client or the valuer, this is a very sound crosscheck.

In addition, if your predominant method, or primary method, is to use an EBITDA multiple, what PE multiple results, or what sales multiple? Can these PE and sales multiples be reconciled with those same multiples for the comparables?

Another method of cross checking your multiples is to perform what's called a first principles analysis. This is essentially doing a one period DCF of the business using a cost of capital, prospective growth rates, and capital investment requirements, which allows you to effectively work out what the multiple would be on a first principles basis. This can often reveal challenges to a valuation using comparables alone.

Sin #6: The ability to change your mind

In finance, the ability to change your mind is embodied in your ability to get out of the investment.?In valuations parlance, these are the concepts of liquidity and marketability.

Oftentimes, discounts are applied for absences of liquidity or restrictions on marketability. These are supported by a wide range of valuation evidence. However, the way in which they are applied is widely divergent amongst valuers.

As in all things, averages are informative, but only if your company or situation matches the circumstances of that average. Accordingly, it's important to look at the risk, the period of your illiquidity, and the cash flows which mitigate your liquidity, in determining what sort of discount might apply. If you ignore these factors, the discounts that you use in your valuation can be easily challenged.

Sin #7: Capital offences

Common ways in which capital related errors arise in multiples are as follows:

Using the same EBITDA multiple for a business with a new capital fleet as for an older one, which has the prospect of significant capital outflows much closer in time

Using multiples adjusted for AASB 16 with earnings that have not or vice versa, and failing to adjust for differences in the extent of leasing across comparables even when using the post AASB16 earnings and implied multiples.

Failing to eliminate the effect of surplus assets or liabilities in the comparable set, distorting the multiples which you then apply to your subject business

Failing to adjust for working capital seasonality differences between your company and the comparable set

When an operating business owns an asset that comparable companies usually lease, the earnings from the lease stream may have a different multiple required than the underlying business earning stream.?The largest issue comes up for property rich trading businesses.

Each of these issues can lead to major valuation errors if not carefully considered.

Take-aways

Multiple valuations are still useful, and very common.

The reason, they're easy to argue, are computationally straightforward and easy to cross-check.

However, like DCFs, they can be prone to error if issues, such as the above, are not addressed.

If you normalise well, apply the multiples to the correct earnings, adjust for the differences between your company and the comparables, crosscheck your answers, take account of the appropriate exit factors, and deal with capital issues appropriately, you can improve your valuation reliability.?

-------------------------------------------------------------------------------------------------------------Richard Stewart OAM is a Corporate Value Advisory partner with PwC.?He has been with them for 35 years in Australia, Europe and the USA, doing his first valuation in 1992. He has helped his clients achieve great outcomes using his value skills in the context of major decisions, M&A, disputes and regulatory matters.?His clients span both the globe and the industry spectrum.?He holds a BEc, MBA, FCA, FCPA, SFFin, FAICD and is an accredited Business Valuation Specialist with CAANZ.?He has written two books, Strategic Value, and Hitting Pay Dirt, and is an Adjunct Professor at UTS. The opinions in this article are his own and not necessarily PwC's.?