This post is directed equally towards 1) entrepreneurs trying to figure out how we value companies and 2) regulators, auditors and such trying to get true market pricing for our assets.

I was recently talking with Eric Nath who is one of the more forward thinking appraisers. We were discussing the mess that all of the new regs and guidances have created in some of the more illiquid asset classes. As discussed in recent posts, the venture industry has an issue on its hands because it is an imprecise world being forced to use precise valuation tools. Many would say that it is about time that we came out of the stone age and accepted more responsible, transparent methods. Furthermore, our industry group, NVCA, has embraced the new valuation recommendations.

So, if we are to join the 21st century on valuation, we need to do two things...determine the discount rate and apply it to our expected cash flows. We should be able to use CAPM to determine our discount rate. However, I don't know of any public segments that enjoy 50% failure rates and 10% wins in the 8-10x range. I am certain that there is a beta hidden in there, but efforts to bake this into Black-Scholes models have resulted in both divergent and surreal results. So, CAPM has generally been ignored for early stage VC. Next, groups have tried to replicate how VC's think and back into rates. Early Stage VC's target 10x return on capital in 3-7 years and Late Stage VC's target 3-5x in 2-3 years. However, you get massively different IRR's depending on your scenario (timing & multiple) that look like:

* Early Stage VC is 10x in 3-7 years

-- at 3 years, this is a 115% IRR

-- at 7 years, this is a 39% IRR

* Late Stage VC is 3-5x in 2-3 years

-- at 2 years with 5x, this is a 124% IRR

-- at 3 years with 3x, this is a 44% IRR

So, if you average these out, you get roughly an 80% IRR for early stage deals and an 84% IRR for late stage deals. This is a pretty high discount rate and a pretty large range for the IRR's. Furthermore, if you ask most Late Stage VC's, few will tell you that they are targeting an 85% IRR. So, I am not certain how you use even our own words to back into the discount rate.

Let's say that we have figured out what discount rate to use (80%???). All we need to do is apply it to our cash flows to get a present value. The good news is that late stage VC's should have enough history and profitability in their deals that they could look out 2-3 years and have some confidence that they are in the ballpark. However, their early stage brethren are not so lucky. I don't know any of our early stage companies that can accurately predict revenue (or CF) more than two quarters out let alone guessing at 2-3 years. Most plans we get from entrepreneurs show massive hockey sticks that rocket to $100m+ in revenue by year 5. Most companies are fighting to get over $20m by year 5. Furthermore, if you formally tracked the accuracy of these five year forecasts, you could drive a truck through them. For the 50% of deals that die, revenue might never get above $2m and for the breakout deal, revenue could be north of $100m or it could get its valuation based on number of members, subscribers, etc but only have revenue below $20m. Either way, the gap is tremendous with year 3 cash flow swinging from -$12m to $15m (or more in either direction). Two companies could look like:

Cash Flow Year 1 Year 2 Year 3

Company One -$10m -$12m -$15m (increasing due to scaling costs but poor revenue)

Company Two -$2m $2m $10m (linear traction and growth)

If you apply any discount rate to Company One, you get a negative value and if you apply 40% versus 120% to Company Two, you get a huge different in value.

So, you see the magnitude of the issue. Even by our own standards, the applied discount rate range is enormous due to uncertainly around the timing of the exit. You can use an average, but the standard deviation is massive and would lead to massive mis-pricing for everything other than in the middle of the bell curve. We also have significant uncertainty around the actual cash flows.

It is possible that I am overlooking some clear analogy that would result in a straight forward methodology. In fact, I would pay significant money to any appraiser or regulatory who could help us get visibility around future financials. I don't enjoy reporting bad news to my partners any more than accountants like imprecise methods. This would save us tremendous heartache post-investing. You have my email above and the comments section below...our phone banks are open and awaiting your calls.

Hey Matt,

I wonder if trying to look at the appropriate discount rate for a valuation is always the correct question to ask in the first place. Why does valuation need to be determined in 21st century via discounted cash flows only? As you rightly point out, this method falls if the forcasting exercise is a high beta one. What I have always used when I write business plans is a valuation matrix that shows the valuation under lots of different methods and a high, medium, and low for each i.e.

.

____________________________Low______Medium___High

Discounted Cash flows_______Value1___Value 2__Value 3

EBITDA Multiple_____________Value4___Value 5__Value 6

Sales Multiple______________Value7___Value 8__Value 9

Average of all methods______ValueX___ValueY___ValueZ

Median of all methods_______ValueQ__ValueR___ValueS

Low Valuation________________________Value 1

High Valuation_______________________Value 2

Etc. (and by the way I typically use a 70%, 50%, and 40% discount rate for the above Low, Medium, and High scenarios for the discounted c-f method)

In so doing we have a more robust picture of what the value might be. I guess you could say I use Meta-Valuation approach. This is helpful, since if for example one thinks the company is likely be sold quickly, it is more likely a buyer will use a multiple method based on other comparables, whereas if the company is felt to require lots of capital in later years and not be a easily flipped, discounted cash flows might be more appropriate. Most importantly, if there is a huge range of valuations, one may want to examine the underlying revenue and expense assumptions.

.

Valuation is always a tricky thing, especially in the early days. When we did our Series A round at LiquidTalk with our VC's, we avoided the topic entirely and just said to our investors "we are willing to dilute by X%; you can decide how much capital to put in and that will in effect set the valuation". Kind of a backwards way to do it, but avoided alot of debate about valuation and it worked as we were generally in the same ballpark about what range the business needed to get to the next level.

Hope this is helpful.

d.

Posted by: Dave Peak | April 26, 2007 at 01:00 PM

Matt,

You might want to check out the area of valuation known as Real Option Theory. I think this is where the industry is headed for this type of analysis. It's the only thing I have seen that makes sense of the VC style of investing...

Posted by: Alex Hodgkin | June 05, 2007 at 06:30 PM