Valuation in the real world

Each semester, when I talk to MBA students about finance, one of the big themes is how the practice of valuation in business differs from what is taught at University. The short answer is that it differs a lot, but probably not in the way you expect. The following is a summary of some of the issues that come up in these talks.

Even the most experienced valuer might be surprised to reflect on some of these differences, and I’d be interested to hear any readers experiences in the comments to this article.

The problem with being a professional valuer

Apparently, 80% of drivers think that they are better than average drivers. Of course, this couldn’t be true, as if driving ability is normally distributed 50% of drivers are worse than the average and 50% above.

In my experience, the stats are even worse for business valuers. Pretty much everyone thinks that they are better than the average valuer, especially on the most difficult valuation problems, like early stage or high tech businesses.

Curiously, this might be an example of the Dunning-Kruger effect, where the least competent people are the most confident of their abilities, because in some areas the practice of valuation can elude even very experienced and qualified valuers.

Nevertheless, the layman’s high perception of their valuation proficiency creates a major communication issue - how do you illustrate the complexity of the task and the soundness of your analysis without losing the business, seeming like an egghead, or worse, failing to convince your audience of your conclusions?

Finance theory is just a start

Uni students get a good education in the finance theory and mechanics behind concepts like discounted cash flow or residual income valuation techniques. They might even get an intro to data tools like CapitalIQ or Bloomberg that eliminate much of the research legwork that used to soak up a valuer’s time.

That is a great start, but far from what informs a valuation in the real world. For a start, most uni students prepare their own estimate of the value of the business, and consider that a valuation.   This process requires an understanding of strategy, finance, tax, and financial economics, but is it a valuation? Regrettably not; as formal valuations are typically estimations of the price that an asset might trade at, rather than an individual’s estimate of value.  

On top of the finance theory and individual value estimates an understanding of the negotiation position and likely tactics of the parties is critical to estimating value. For example, in a real estate auction, every individual has their own estimate of value based on their assessments of the attractiveness of the property. These individual assessments are akin to the “valuations” prepared by Uni students. A professional valuation is an estimate of where the price will settle - which is usually at the value just every so slightly higher than individual assessment of the second highest bidder, at least in this simple example. This is the difference between price and value that is well known to any long term investor.

Psychological biases also have a role to play where they reflect how behavioural economics interplays with classical economics in markets. Herd mentality, risk aversion and other human factors play into the way market prices move with just as much impact as your perfectly formed DCF analysis.

There are not just knowledge gaps. Valuation is not a fact, it is an opinion, based on judgements. The better your judgements are supported by logical arguments and compelling evidence (all wrapped in a compelling narrative) the more likely your opinion will stand. So communication, investigation/research skills and the principles of logical argument are equally important skills as finance ones for the professional valuer

Your model might be good but it is only a model

It’s quite natural to be proud of a lot of hard work. Most models are just that; often almost a labour of love for some. But models are models, which by definition are simplified (even if often very helpful) views of complex real world phenomena. 

Because of the simplifications made in models, it is important to be aware of where the model may not fully describe reality.

The most vivid example of this was loan portfolio valuation models prepared in the run up to 2008. One of the simplifying assumptions was that loan losses were independent of one another. It turned out that when economic and financial market turned south, this assumption failed, leading to the Global Financial Crisis. Whilst not all valuation models feature as dramatic issues as this, they all have the potential to do so. Stress testing your conclusions to large shifts in factors around the business will often reveal these sort of issues.

Build a toolkit

There are three general valuation approaches; cost, income or market. Whilst each of these approaches describe how owner of the asset can extract value (by avoiding costs that have already been incurred, or enjoying using the earning capacity of the asset or through sale), within each of these approach categories there are a panoply of techniques like depreciated optimised replacement cost, discounted cash flow, residual income, dividend discount model, real options, scenario analysis, regression of comparable companies and so on.

One of the key skills for a valuer is choosing the approach and related technique to employ, which requires consideration of, at least, the following:

• The nature of the asset (eg classic cars or residential dwellings lend themselves to market approaches rather than income ones)
• The availability and reliability of various information sources
• The consequence of the valuation outcome
• The contestability of valuation conclusions
• The relevance of the approach or technique to the audience of the report, and
• Risk appetite for the valuer

Having as many tools in your valuation technique tool bag will allow you the flexibility to deal with these issues as they emerge in your work. Because of this need for adaptability, this makes the use of AI for valuations limited to specific asset classes and valuation circumstances where information is both widely available and reliable, at least for the moment. Probably good for the longevity of a broadly based profession; but it does mean you will need to continually invest in your toolkit and deliberately answer the contextual questions that result in the choice of methodology.

It's about the data, man

At University, valuations focus on the calculations, and rightly so. This means that the data is usually provided in nice easily accessible packages. 

In valuation practice, often these numbers are less easily acquired. Consider, for example, valuing a business that is to be carved out of a larger one. The valuation process will feature a large amount of construction of pro-forma financials and projection of those proformas consuming often very much more time than the DCF analysis.

On a brighter note, luckily information providers have rapidly sped up the search for and analysis of comparable listed companies and transactions, reducing the burden of this sort of information acquisition for valuers. The counterpoint to this bright spot is though that much more analysis of the basis of the comparison with the subject company is now both feasible and expected.

How not to be wrong

Valuations are based on judgement, and the strength of your arguments. Accordingly, there are a wide range of plausible valuations, and they will often differ between valuers, even if they have used similar methodologies. The two ways you can be “wrong” are being unreasonable or making a manifest error. The easiest one to find is a manifest error, so we’ll tackle that one first.

Manifest errors largely arise in two ways:

• Calculation errors (in your model or report)
• Conceptual errors (like applying a forecast multiple to historical earnings)

These two issues can blow your valuation out of the water, but because they are relatively easy to find, peer review can usually sort them out quickly.

Being unreasonable is a far harder case for anyone to prove, but they are many more ways in which they can arise. Here are a few examples, with their remedies:

• Unreasonable or unsupported assumptions.  This can happen in any valuation technique, but each assumption requires evidence and rationalisation. The less supported the assumption the more likely someone will claim your valuation unreasonable. These can be minimised by benchmarking key assumptions to guideline companies or other valuers.
• Unreasonable outcomes when individual inputs appear reasonable. This most often happens when using DCF models. Because of the multiplicative nature of a DCF calculation, small variations in each assumption (within the range of reasonableness) can lead to a highly unreasonable outcome. The only real way to minimise the risk of this one is to perform cross checks using multiple valuation techniques.
• Non-normalised financials. Businesses that have very large one offs events, non commercial arrangements, or changes in accounting, can have events which don’t reflect a realistic expectation of the future. A bit of diligence about these issues, plus some benchmarking will find and remedy the potential for this cause of unreasonableness.
• The incredible growing business. This usually arises in terminal values, where even small growth above inflation will mean that the business is forecast to eventually take over the world. This is due to the impact of exponential growth over long periods of time. The way to avoid this one is to be very careful about what assumption you use for terminal growth.
• Ignoring mean reversion. Think of this one as Newton’s law of gravity for finance. Mean reversion means the tendency for unusual variations in data to tend towards an long term average. Accordingly, high growth businesses tend to return to average in the longer term, and high profit businesses tend to revert to average profitability (there are many examples). The forces driving this return is not gravity, but competition and the law of large numbers. Reviewing your assumptions for their tendency to revert to average is a way to check for this issue
• Simplistic analysis. This is usually where a whole bunch of data is gathered and the valuer chooses the average, or an arbitrary cut-off. It usually happens in estimation of premiums and discounts, or the use of multiples. Understanding the range of outcomes in any assessment can highlight a driver the valuer should understand in order to calculate an appropriate value. Graphing the relationship between your numbers and potential explanatory variables will help valuers clarify this issue.

Dealing with luck in finance

Hindsight is a wonderful thing, but it makes valuers look a little silly all the time. Imagine you say something is worth a dollar today, and in a year’s time it’s worth two dollars. If you bought it, you look like a genius, but if it fell to zero, you look like a mug.

If both outcomes were equally probable, the valuer was right today, particularly if the everyone was informed about the risks and the market was liquid. Unfortunately, that doesn’t mean they won’t be criticised if the price hits the floor a year later.

This is where luck (and good decisions) come into finance. If you paid $1.50 for the asset in our little example, and the price went up, you feel like you made a good decision. You didn’t: you just got lucky. The reverse is true, if you paid 50c and the price tanked.

The lessons - be sceptical of someone saying that they are a genius with a small number of successful investments. And be very sure to highlight in your reports the risk that the future could look different to your valuation.

Conclusions

While understanding these issues will not make Uni students or laymen expert valuers, they may help them deal with some of the pitfalls that drive overconfidence and bad investment decisions. 

A side benefit might be a greater understanding of the dimensions of valuation decisions and the value that an expert valuer can provide to help them through those types of decisions.

-------------------------------------------------------------------------------------------------------------

Richard Stewart OAM is a Corporate Value Advisory partner with PwC. He has been with them for 33 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 (2012),and Hitting Pay Dirt (2017). He is also an Adjunct Professor at UTS.