Last week Fred Wilson and I had a good back-and-forth going via comments on his post about venture capital returns . I’ve been thinking more about what I wrote concerning the proper measurement period for calculating returns and I thought I should follow up to Fred’s last reply . So, here are a couple of thoughts: (1) No matter how you calculate returns, an actual improvement will help the industry. The internal revenue code (IRC) is biased against venture capital investment in a material way. One change in the IRC would have a material impact on venture returns. More below. (2) I’m no accountant, but I believe LPs in venture funds have to account for their investments as a liability from the time they agree to invest. (3) Maybe, LPs in VC funds have sophisticated models anticipating calls etc. But that may be more scary than good.

IRS should revise NOL rules

One of my partners and I recently pointed out in an article in VCJ that the current treatment of net operating losses (NOLs) under the internal revenue code shows a material bias against the venture investment model. Here is the gist of that article:

Venture-financed companies create NOLs by spending the proceeds of financings to develop and grow their businesses to the point of exit. If the company is successful, one of two outcomes typically follows: either the company may be sold, or the company may have an IPO. Under current law, such a sale, and in most cases also an IPO, will cause an ownership change that will greatly diminish the company’s ability to use its built-up NOLs. This is particularly true if the company has had one or more financings within three years before the sale or IPO.

If the company were to continue to run its business without an exit, however, the company could freely use its built-up NOLs to shield profits from taxation. Venture investors may be faced with two choices. They may be able to sell (with the consequent diminution in NOL availability) or continue to operate the company until it has utilized its NOLs and then sell. The built-up NOLs represent a significant tax savings that can be realized only if the company is not sold. [the tax savings can roughly equal the amount invested in the company multiplied by the applicable corporate tax rate.] As a result, current tax law shows a bias against venture capital investment. A similar, but somewhat more complex analysis, applies to IPOs that result in an “ownership change” and a consequent limitation on NOL utilization.

Removing this restriction on NOL utilization would greatly facilitate exits, because venture investors could reach their return rates more quickly and because these transactions would be more attractive to acquirers. A company valuation -- whether in a sale or an IPO -- should be higher with fully utilizable NOLs.

LP Accounting

I am not an accountant, but my understanding is that accounting rules that apply to LPs in VC funds require them to carry their commitments as a liability on their books to the full amount of the commitment. If that is the case, shouldn’t they have to measure the return from the time they incur the liability?

LP Models for calls

I take Fred Wilson’s point to be that LPs have many many investments that can be called, and that LPs have developed statistical models to predict how much cash they need to keep on hand to cover potential calls. The LPs are nonetheless liable for all those calls. They are merely gambling that not all the calls will come in at the same time. In effect, they are pooling the commitments and hoping that only a portion will be called in any given time period. Shades of other pooled risks.