How to better validate reason for private-company discounts
Editor’s note: In the Q4 2016 issue of Business Valuation Australia, Gilbert Matthews wrote about whether private companies are valued for less than public companies solely on the basis of being private or whether other factors are involved. In this issue of BVWire Australia, Tony Carlton (Macquarie University’s Applied Finance Centre) shares his viewpoint.
Gilbert Matthews refers to research by Micah Officer (2007) that found that, compared to acquired public companies, private companies sold at average discounts of 17% but that subsidiaries (of listed parents) sold at average discounts of 28%. Matthews suggests this result should be subjected to further empirical research.
Like Matthews, I was uneasy about this result, particularly the fact that unlisted subsidiaries had reported discounts much higher than private companies. The research to which Matthews refers attributed the discounts to parent companies being under financial pressure and thus selling at a discount—effectively the fire sale argument. There is good evidence (and logic) that, most commonly, public companies are likely to exit underperforming or noncore assets. Spinoffs or IPOs have the potential to exit assets with high profitability and/or growth potential. Assets sold by trade sale are therefore more likely to have lower underlying valuations than public market comparables. Consequently, any reported discounts will simply reflect differences in underlying values rather than any form of private-company discount.
Using U.S. data for the period 1997-2009, I compared transaction multiples for the sale of unlisted subsidiaries by public companies with those for the acquisition of comparable publicly listed companies. Comparable acquisitions were selected based on time period, industry, and size. I found a small subset of fire sale-type transactions for which there was a material discount to comparable transactions. However, the number of such transactions did not explain the presence of pervasive discounts.
More importantly, I found that the discounts Officer reported were largely attributable to the measurement methods used. There are three key decisions to be made in measuring these discounts:
- How you calculate the average of the comparables (arithmetic, geometric, harmonic means, or median);
- How you calculate the discount (percentage or logarithmic); and
- How you select the sample (particularly outliers).
Officer used the arithmetic mean, percentage discount, and eliminated a number of outlier transactions. This combination explains the high reported discount and also makes the results very sensitive to the treatment of outliers. Using alternative averaging methods and calculating the discount using the logarithmic method resulted in a significant reduction in reported discounts. In fact, the samplewide discount was lower than that for private-company sales, more in line with intuition (Note: In some cases, the discounts were eliminated). Furthermore, the results were much more stable to different treatments of outliers.
I then tested whether asset-specific characteristics help explain the discount differences. As an alternative to using multiples, I constructed a regression model to calculate the “warranted value” of a deal, based on size, profitability, and industry type of each target. This approach allows a target-specific estimate of fair value using each target’s own profitability. Using this method, reported discounts averaged zero for the sample. This implies that asset-specific characteristics largely explained reported discounts, in line with Matthews’ suspicions. Other analyses further confirmed the role that asset characteristics, particularly profitability, play in explaining variations in transaction values. These results were further confirmed by analysing share market reactions to asset sales, where I found no evidence that the market reacted as if unlisted subsidiaries were sold at discount to underlying value.
My conclusions from this research are threefold. First, discount calculations are sensitive to the measurement procedures used. The combination of arithmetic mean and percentage discounts generates results that are sensitive to the sample and the treatment of outliers. At a minimum, analysts should test the robustness of their results with sensitivity analysis to alternative methods. Second, it is important to allow for asset-specific characteristics, particularly size and profitability. This can be achieved by appropriate selection of comparable transactions or the use of regression-type approaches as an alternative to the simple multiples approach. Finally, this research highlights the need for the reporting of appropriate financial data on private transactions. Financial data on private trade sales are not as available as they are for public companies, and much of the multiples analysis ends up based largely around revenue multiples. If we are really going to decide whether a transaction is at fair value, we need to be able to use target-specific financials to form such a view.
To read Gilbert Matthews’ original assessment of discounts for private companies, activate your subscription to Business Valuation Australia today.
back to top
Valuing startup businesses—the riskiest work you’ll do as a valuer
Startup business valuations are challenging for many reasons, but a big challenge for valuers is the risk inherent in preparing the valuation itself. This is because of the nature of the businesses; there’s some probability that the valued business will be worth a lot, but a much greater probability that it will be worth a very small amount. This means that, about nine times out of 10, with hindsight, someone will say, “Well, this business was overvalued.” Those same people won’t thank you the one time out of 10 that the business becomes worth 10 times the valuation amount.
The reason it’s risky is that, with hindsight, a valuation for a startup almost always looks overcooked. Valuers need to be aware of this risk and manage it through the engagement contract and the communication of the results.
Read more about the challenges to valuing a startup company and how to manage the risk in Richard Stewart’s article, “Valuing Startup Businesses—The Riskiest Work You’ll Do as a Valuer” in the Q4 Business Valuation Australia.
back to top
New IVSC board looks to emerging markets
The International Valuation Standards Council (IVSC) recently formed a Membership and Standards Recognition Board that will focus much of its efforts on the IVSC's objective to be seen and referred to as the global standard setter for valuation with international valuation standards. Among other activities, this board will play a key role in helping respond to requests to develop the profession in emerging markets. The new board’s chair is Eric Teo (IVAS, Singapore), and its members are: Aart Hordijk (NRVT, Netherlands); Allan Beatty (AIC, Canada); Eleanor Joy (CICBV, Canada); George Badescu (ANEVAR, Romania); Jeannette Koger (AICPA, U.S.); Jiang Wei (CAS, China); Ken Creighton (RICS, Global); Phil Western (API, Australia); and Shigeko Mizutani (JAREA, Japan).
back to top
Brand values for large home appliances
In some industry sectors, the branding strategy is the most important element of competitive strategy. One such example is the home appliance sector, where profitability depends a lot on the branding strategy. While large parts of this sector follow a mass market and large-scale production approach, some players follow a niche and premium strategy.
Mature market: This month’s peer group analysis from MARKABLES examines brand and enterprise value multiples for 26 home appliance businesses (“white goods”) acquired between 2001 and 2015. The peer group includes brands such as Indesit, Hotpoint, Sanyo, Enodis, Amana, ProLine, US Craftsmaster, GSW, ATAG, and others. Operating in a rather mature and competitive market, these businesses are valued at an average of 1.0 times revenues. Brand accounts for approximately 25% of enterprise value, and the median trademark royalty rate is between 2% and 2.5% on revenues. This is not really impressive considering the brand awareness and marketing spend in this sector.
A few brands stand out from the pack, however. These are niche brands following a premium strategy, such as premium cooking ranges, wine coolers, or luxury outdoor grills. Despite their disadvantages in scale, they show higher profitability and brand value multiples. Their trademark royalty rates are above 5%, and the brand/enterprise value ratio is 35% and above. Not surprisingly, profitability of these businesses is twice as high, with enterprise value at 2.0 times revenues. Premium branding can help to increase value, but it is restricted to the smaller volume in the niche.
MARKABLES (Switzerland) has a database of over 8,200 global trademark valuations published in financial reporting documents of listed companies.
(Click image to view full size)
back to top
Should you do site visits to your subject company’s competitors?
In some industries, visiting your subject company’s competitors is very important. For example, visiting competitors is a must when valuing a hotel, according to Mark Dayman, an expert on hotel value and vice president of FranIntel, a franchise consulting company. “Suddenly the quantitative information you’ve developed begins to come to life,” he says. “You start to understand why your client performs better than the others within that market. You also learn what is missing in the market that may be an opportunity for your subject entity. Those interviews are really important.” If you can’t see top management, try talking to other key people. For hotels, Dayman likes to talk with the front desk managers.
Extra: For an in-depth look at valuation in the lodging industry, What It’s Worth: Hotel Business Value is a must read.
back to top
