Most business appraisal assignments are for small private companies with revenue less than $10 million, yet current cost of capital estimation methods rely almost entirely on large public company security returns. But these small private companies differ from large public companies in so many fundamental ways, and consequently, there are issues that make current methods for developing discount rates unreliable when they are applied to small enterprises.
IPCPL is steeped in market transactions of small private businesses … think of IPCPL as the security market line of private company transaction data. It was developed to eliminate the problematic issues we face (e.g., company specific risk premium, tax rates for pass-through entities, leverage, liquidity discount) and to create a more defensible starting point for deriving the cost of capital of any private company with revenue less than $150 million.
Welcome to what I think is an interesting topic – and probably even a controversial one at that.
*** The Implied Company Pricing Line & Model ***
Maybe you’ve already read something about this … watched one of our webinars … or seen a previous presentation.
Ask for show of hands if you’ve watched one of our webinars or read on of our papers.
No matter how you feel or how much you know about the topic, I hope you will leave here today with some useful information.
A rogues gallery if I’ve ever seen one!
So, why ARE we here today? Sometimes, we feel this way!
And we’re here to annoy you … about the BUM and a few other generally accepted BV practices.
I used to hate testifying about how I developed my discount rate.
First, you start with the Rf rate … then you add this … then you add that.
No one thinks that way about a rate of return – accept academics.
In the land of the blind, the one eyed man is king.Erasmus of Rotterdam, circa 1510
Caveat: Could use for minority valuations … with minority CFs … but would need incremental/additional adjustment for DLOM
For y’all here today, I’ve got 4 learning objectives.
The pitfalls are part and parcel of “generally accepted BV practices.”
And, frankly, are problems for most BV assignments.
Worse, we know the following issues to be true.
Chart from Gary Truman
Duff & Phelps would agree
So it’s really NOT the cost of equity as if our subject company were publicly traded.
We need to adjust the numbers for unsystematic risk, illiquidity and, perhaps, tax effects.
If the SSP exists, why stop extrapolating at 12%?
Is the SSP capturing something else, e.g., illiquidity?
Chart prepared/created by Toby Tatum
Note: 10z Size Premium is 12.12% at 12/31/13 and 11.98% at 12/31/14.
For Rod: Log - 4 = $10,000 ; 6 = $1 million; 8 = $100 million; 10 = $10 billion; 12 = too big to fail!
Both cases are from Delaware Chancery Court – Open MRI, April 2006; Gesoff, May 2006.
And to be fair, there have been many cases that have accepted a subjective CSRP.
Damodaran on illiquidity …
Also, this: The trading costs associated with buying and selling a private business can range from substantial to prohibitive, depending upon the size of the business, the composition of its assets and its profitability. There are relatively few potential buyers and the search costs (associated with finding these buyers) will be high. In fact, if the investor buying it from you builds in a similar estimate of transaction costs she will face when she sells it, the value of the asset today should reflect the expected value of all future transactions cost to all future holders of the asset.
Two BUM supporters; much different conclusions for a control valueOur ongoing surveys of BUM supporters prove unreliability
With all of the previous issues – there are no “how to” instructions for dealing with a control valuation.
IPCPL is agnostic on these issues – it’s ALL baked into the private company transaction data.
Regression points for $1-20 million revenue and every $5 million after that.
Probability neutral - just as much chance things will turn out better/worse than this forecast.
Two different company forecasts may have equal chance of occurring, but one could have a higher std dev require higher discount rate.
Note: Risk adjusting is modeling the CSR in the CF forecasts.
Tests:
Pepperdine: COE is 22% on deals levered more that 2 to 1.
Hitcher: If you take D&P COE data and merge with 30% debt and cost of debt, you’re close to IPCPL IRR.
BVR: Russian guy testing IPCPL model (Igor ???? – physics and economics background) finds model to be ok, Grabowski one of the reviewers.
Capital structure: Best possible prices have optimal capital structure/cost of debt; with BUM, have to assume a capital structure … or which unlevering/relevering formula to use.
Pratt’s Stats MVIC excludes cash, even in stock deals.
Thus, only add back cash that can’t earn interest, e.g., cash in cash registers.
All available dates? – same as new BVR paper
No medical/dental practices (compensation [separating business value & labor value] and financial report issues), but IPCPL can be used to value them? – S&P 500 has utilities in it, but we don’t value utilities.
Why only private buyers of private companies? – see later slide
Just like assembling a portfolio of diversified stocks reduces risk, the more transactions the lower the potential error created by any errors/noise in the data.
Ex. – given a starting Ko of 20%, a 1% increase in g would result in new Ko of 19.5%.
Not a 1 for 1 effect because increased g requires higher reinvestment in NWC and Capex.
Net result of adjustments not a material issue
Present Day – 20 years … inflation different … ERP different change to “today”
Also note: stock sales v. asset sales – not a statistically significant difference
Stock sales occur when g is slightly higher, so Price/Sales slightly higher
Updated May 11, 2015
Updated May 11, 2015
$150 million revenue
Cost of capital determined using 3 factor Fama-French model
Note that average company has no net debt
Adjusted for costs of going/staying public
Rule asserts that two securities that provide the same future cash flow and have the same level of risk must sell for the same price.
10% of the first million dollars involved in the transaction
8% of the second million
6% of the third million
4% of the fourth million
2% of everything thereafter
Regression points for $1-20 million revenue and every $5 million after that.
Output updated May 11, 2015
Practical example – IF your company is average
If you don’t like Pratt’s Stats or are uncomfortable with IPCPL … IBs/Mercer use 5x ebitda multiple rule of thumb
Pepperdine 22% COE with debt = IPCPL results
Toby Private Enterprise Pricing Model – uses BizComps … separately for each major SIC code
We realize that IPCPL may be “out there” from some of you.
That’s why we created IPCPM, which is a calibration tool you can use as a sanity check on your BUM.
We put calibrator output in our reports
What is average? Your idea is different than mine … which makes results subjective.
Thus, the calibrator is like having a cost of capital expert helping you in the background to avoid the BUM pitfalls.
Note that there is a downloadable manual on the website.
We don’t know how to move off the curve – to extent we’re not average, we have to guess.
But not quite similar to how we guess at CSRP.
With IPCPL, we are starting in the realm of small private companies so these adjustments are not a hug leap of faith.
But with BUM, we are starting with large public companies … who knows what the adjustment should be?
Factors are percentiles
Note: Avg IPCPL company (“5”) has beta of 1.25
The systematic risk and unsystematic risk factors assume a 50-50 wt between beta and total beta (at the smaller sizes and escalates to a proportionately higher beta wt as size increases) - so just like TB this is a more scientific approach.
Factors are percentiles
Note: Avg IPCPL company (“5”) has total beta of 3.38
Assume marginal (price-setting) investor purchase of private company will represent ½ of net worth
The systematic risk and unsystematic risk factors assume a 50-50 wt between beta and total beta (at the smaller sizes and escalates to a proportionately higher beta wt as size increases) - so just like TB this is a more scientific approach.
Factors are percentiles
Note: Avg IPCPL company (“5”) has illiquidity of 0
Factors are percentiles
Note: Avg IPCPL company (“5”) has debt/capital ratio of 0.33
With all of these adjustments:
It's judgmental based on the dispersion of the transaction data and our experience with M&A - We do not want to say it's anything more scientific than that. IPCPM was developed for people who wanted to be able to subjectively move off the curve with our help.
We have experimented with the percentile inputs and calibrated them to get reasonable results based on the data and our experience in M&A - it's nothing more scientific than that.
Greater than +10% too high
Within +/- 10% acceptable
Less than -10% too low
Greater than +10% too high
Within +/- 10% acceptable
Less than -10% too low
Greater than +10% too high
Within +/- 10% acceptable
Less than -10% too low
If you do what you’ve always done, you’ll get what you’ve always gotten.
IPCPL: One answer for ALLIPCPM: Deviation much tighterBUM: Deviation could be 2x