Question: How do you guys think about valuation multiples? Assume that you’ve built out your model, done all your due diligence, and have a set of comps and precedent transaction. Now, how do you think about whether you want to pay 10.1x P/E for a company, or 10.2x, or 10.5x? I think about it in terms of returns, and I guess to me, it’s a matter of understanding what your hurdle rate/required rate of return was. Once you know what your hurdle rate is, you can potentially back into the multiple that you’re willing to pay. This seems pretty intuitive, but I don’t know if this is the best way to answer this question, or if there’s more to it than I’m realizing. How would you answer this question in an interview?
I would suggest that multiples are relative. you start at the mean of your universe and then add or subtract subjective multiples based on the company’s strategy (therefore growth prospects), margins, the capex requirements (or just right on down to free cash), fx exposure, product lines and so forth relative to the other comps you picked out. at the end of the day, you should always translate what your dcf says into mulptiles and compare relative to the universe…just in case a ‘reasonable’ DCF value translates into an unreasonable multiple…
How do you determine what’s “reasonable”? All I’m saying is, assume for example that you truly feel the company is better than the rest of its peers and should warrant a superior valuation multiple. You’ve already done all your research to make your conclusion that it’s better than the peer group and there’s nothing else you need to do in terms of diligence or modeling. So at that point, how do you figure if you’re going to pay 10.1x, 10.2x, or 10.5x?
valuation, like most of finance is as much as art as a science (which you know). so when you know that it should trade higher, you pick a multiple that supports your thesis…so your argument is that the stock will go higher than it is and higher than its peers (you’ve stated direction)…now you pick a multiple to support you thesis that won’t get you laughed at (ahhh, the art part of valuation)… its all for show, cuz in the end buy side could care less about sell side targets. (which we know)
Then what does the buy-side care about? How does a long-term, fundamental investor think about entry valuation? I already know what the sell-side (banking, research) do…I mean, that’s *my* job. I’m just trying to understand if there’s more to it – why pay 10.6x for a company instead of 10.5x? How do you decide what the exact multiple is? Is this all hinged upon your hurdle rate or is there more to it? And how do funds determine what their hurdle rate is? If it makes a difference, assume this is for a private investment (VC/PE/growth equity) rather than a public one. By the way, I’m not trying to be obtuse here – let’s imagine for a moment this question actually came up in an interview, because sometimes imagination = reality. You don’t always get asked the best questions but you still have to figure out how to answer them without going ape on the interviewer.
buy side cares about direction - does the business fundamentals (strategy, markets, products and so forth) support a rising or falling business value. then you have to make the transition to what you’re wondering about - price vs. value. how do you decide your entry point - great question. growth stock you don’t generally worry about a 10.5x vs. a 10.6x cuz you know next year it’ll trade higher if you’re good at picking the stocks direction. value plays are a little more tricky cuz the price. vs. value equation can leave less of a margin of safety. this is when its usefully to reverse engineer market assumptions that are built into the stock about growth rates etc. what makes the cut in terms of return - ones that meet fund objectives (in terms of diversification, expected return (or IRR, hurdle rates etc.), and ones that meet the funds mandate. of course i’ve oversimplified my experiences with these but i think it’s easy to paint a general picture that way… and i wasn’t assuming you were being obtuse. genuine until proven otherwise.
Hi Numi, I’ve thought about these multiples a little more from the value-growth spectrum, but there are a few ways to look at it. If you have an industry that has more than a few companies, you can look at the industry average multiple and then look at the standard deviation of the distribution of that multiple for all companies. From there you could pick a number from 1 to 99 that represents your estimate of what percentile of strength that company would be in on your subjective basis, with 50 as average for the industry. Look up the cumulative probability distribution for that number (assuming 100 is high), get a z score for that, and then calculate a multiple base on AVG + Z * SD. Now, you’d have to take that number with a grain of salt, and it’s statistical overkill to take the number too too seriously, but it might not be a bad way to turn a gut feel into a target number. The other thing you can do is try to think about what a high multiple represents. As far as I can tell, a high multiple means one of two things. 1) There is a high growth opportunity (franchise value) that is not clearly shown in current earnings/operations figures; or 2) there is a lower level of risk for the company, and this warrants a price premium (and consequently lower earnings yield for less risk). If (1) or (2) doesn’t justify the high multiple, the only remaining conclusion I can think of is (3) it’s overpriced, so sell it. As I dig deeper into this, I’ve come more to the conclusion that it’s a good idea to think in terms of earnings yeilds (reciprocal of P/E), at least if PE is your preferred multiple. Maybe a private company should have an ROA like figure, say, reciprocal of EV/EBITDA. Earnings yields allow you to put negative earnings companies and zero earnings companies on the same scale as companies with positive PE. It gets a little trickier to do things like that with PE. For example, what is the PE of a company with zero earnings? Does that mean that it is never advisable to buy a company that has zero earnings? The main advantage of P/E over Earnings yield (as far as I can tell), is that it is easier to see whether a company seems expensive or not, provided that you are comparing the company to others with a similar growth potential and risk profile.
Thanks for the thoughts guys. bchadwick, we assume that we’ve done all of the analysis you mentioned already. When you say a company should be valued higher than its peers, then the question becomes, “how much higher?” So then how do you quantify exactly how much? These are valuation questions for private companies (not publicly traded equity if that makes any difference). I guess the person was insistent that you could put an exact number on the multiple…do you guys believe this is the case, and if so, what guides your decision to say 10.6x instead of 10.5x? Maybe the question is still as unclear to you as it is to me. Basically, what I’m asking is if there’s a way to pinpoint what you’re willing to pay to buy a company (again, let’s take a PE/VC situation). I had a conversation with someone recently and he/she was insistent that your multiple, assuming you’ve done all your fundamental research, hinges on your hurdle rate. And then when they ask you how you come up with your hurdle rate, I guess it just depends on the risk appetite in your portfolio, or something like that…
Numi, I think you’re on the right track with the though process… There isn’t a one hard and fast answer to your question basically because so much in valuation and finance is subjective. And as a result you get these posts that point out generalities to look for in certain situations. I think at the end of the day, you pick roughly the maximum price. is that price 10.5x or 10.6x…great question without a real answer. I believe at the end of the day, you want a certain margin of safety (i.e. expected return) and you buy up to that minimum return level. how does one determine minimum return necessary for a purchase? no one answer can cover that off…
It’s worth what someone will pay for it. Multiples don’t drive sales, buyers do. If everyone agreed on the price (thus, multiple) for assets… hmm, what kind of world would that be?
Ah, I misunderstood, Numi. If you are investing in growth, then 10.5 and 10.6 may not make a huge difference, since most people know that growth estimates are “crystal ball” type work, so you need to make sure that you will still make money even if you are slightly below your “minimum growth assumption.” It’s the growth investor’s equivalent of a margin of safety in the value world. Mostly you would want to stress test your growth assumptions and make sure that 10.5 and 10.6 both enough money to cover 1) your costs to monitor the investment (remember sunk costs don’t count), and b) your opportunity costs of having your capital tied up in this company rather than free to invest in another. So one way to do this is to have a 95% or 99% confidence range for the justifiable PE for the company, and calculate the expected rate of return for the high value low value and expected value. Then you show that a) the worst case scenario still passes your hurdle rate, b) that your assumptions are plausible, and c) you haven’t assumed away any genuine risks (and preferably have a method to address those risks). How do you get your hurdle rate? I suspect that this comes out to be based on operating expenses (the time it takes to monitor), opportunity costs (given limited capacity to monitor and invest, make sure that your investments are the ones with the highest risk adjusted rates of return), and finally some consideration that the portfolio of company investments needs to have an expected rate of return and volatility that the investment policy has promised the customer. So, you might be willing to accept a lower company hurdle rate if the other comapnies in your portfolio are appropriately diversified, but if this bet is highly correlated with existing portfolio holdings, you’ll need a higher hurdle rate. My understanding of value-land is that PEs tend to be compared more to industry averages, and you simply buy the lower and sell the higher, with the caveat that you need to make sure that the low value doesn’t indicate that something is seriously wrong with the comapny (usually some big, highly skewed tail risk that that has some leading indicator). So it’s basically looking at whether a company is out of favor for a justified or unjustified reason. I suspect that most quant arbitrage models just assume that the majority of mispricings are unjustified and will pay off because the justified mispricings (relative to the model) will be a very small portion of the trades … they then get fried if there is some systemic issue that means there are a large number of justifiable pricings that show up as mispricings on their model. If you are in value-land, a PE of 10.5 and 10.6 may make a much bigger difference, because mispricings are likely to be smaller on average (there is still some, but less crystal ball work in value-land). You’d want to see how your range of estimates of PE compare to the average size of mispricings that your portfolio management model accepts. [how do I know this stuff? - mostly because my GF is a growth manager that uses GARP, which is basically a value methodology applied to a growth universe, and we talk about this fairly frequently] Finally, you might just have a regression that looks at company returns and volatilities as a function of your PE or other multiple, the trick here would be to make sure that appropriate control variables are also in the regression (market beta, industry, probably size, etc.)
I think there’s a difference in determining a multiple to when you are buying stocks or negotiating a buyout. In the first situation, the investor is generally a price taker. You know the price you have to pay, the question is rather you think the price is reasonable. So it makes sense in this situation to base the entry multiple with comparable companies. In the second situation (Buyouts/PE/VC), you are actually at a negotiation table discussing the price with the seller/management team, no longer a price taker. There may be a multiple that is reasonable in relative to the comp companies, but the investment still doesn’t make sense (and thus will never be approved by the investment committee) if it doesn’t meet the fund’s IRR hurdle. The comp multiple also does not capture various negotiable clauses (dividend preference, milestone payment, purchase price clawbacks), control premium, leverage, and synergy with other portfolio companies. This is why a case can be made to look at the entry multiple in terms of the IRR hurdle of the fund.
in my experience, multiples are just a guide to see if what you are paying is reasonable. nobody gives a hoot whether you pay 10.5x or 10.6x. Like what your friend said numi, it’s a by-product of your hurdle rate. To get 25% IRR for an LBO, you pay what you have to pay…
Zuran Wrote: ------------------------------------------------------- > I think there’s a difference in determining a > multiple to when you are buying stocks or > negotiating a buyout. > > In the first situation, the investor is generally > a price taker. You know the price you have to > pay, the question is rather you think the price is > reasonable. So it makes sense in this situation > to base the entry multiple with comparable > companies. > > In the second situation (Buyouts/PE/VC), you are > actually at a negotiation table discussing the > price with the seller/management team, no longer a > price taker. There may be a multiple that is > reasonable in relative to the comp companies, but > the investment still doesn’t make sense (and thus > will never be approved by the investment > committee) if it doesn’t meet the fund’s IRR > hurdle. The comp multiple also does not capture > various negotiable clauses (dividend preference, > milestone payment, purchase price clawbacks), > control premium, leverage, and synergy with other > portfolio companies. This is why a case can be > made to look at the entry multiple in terms of the > IRR hurdle of the fund. thanks Zuran – super response.