Vintage Year

You Get the Idea

By convention, a private equity partnership's "vintage year" typically is the calendar year in which it accepts commitments from limited partners. For example, four of the 12 underlying funds employed by TIFF Partners I have 1996 "vintage years"; the remainder were corked in 1997. Actually, "corked" is the wrong verb to use in this context. Unlike bottles of wine, which are filled at a single point in time and consumed at a single future point in time, most PI partnerships have phased beginnings and ends. More specifically, they are offered to prospective investors over a multi-month time period; they conduct one or perhaps several closings toward the end of this period, with each closing entailing the actual drawing down of just a fraction of total committed capital; and they draw down the remaining committed capital over the subsequent one to five years. They then operate for five to 12 years, returning (if all goes well) a combination of cash and securities to their investors. Depending on covenants contained in partnership agreements, the General Partners of PI vehicles will typically raise follow-on funds before 100% of the capital committed to prior funds has been invested. In other words, the gap between the vintage years for consecutive funds in the series offered by a typical PI advisor is less than the funds' planned lives.

Relevance to TIFF Members?

Pinpointing the vintage years of the various partnerships managed by a prospective advisor is crucial in making an intelligent assessment of the advisor's past track record and hence also of the advisor's future potential. Contrary to the belief that underlies some institutions' efforts to cram as much money as possible into PI's, some PI's do not generate huge returns (!). They do not do so for a host of reasons, most of which are no different than those which cause disappointment for marketable securities investors: overall market conditions can cause entry prices for all equities (public or private) to be inflated � or exit prices to be depressed; investor enthusiasm for certain technologies can cause too much capital to flow into a given sector (e.g., wind turbines in the late 1970s, PC disk drives in 1982-1983); and investor anathema toward particular industries can cause exit prices to be unduly depressed when a PI fund is winding up its affairs (e.g., healthcare funds in 1994). Accordingly, institutions that seek to gauge whether a prospective PI advisor has been skillful or lucky must take care to compare "apples to apples," i.e., to assess the advisor's track record in light of the specific vintage years of each of its various funds. These same institutions should also recognize that, as with marketable securities managers, truly exceptional track records are almost always the product of both skill and luck. Indeed, as a general rule, what separates highly successful investment managers from truly legendary ones is not the latter's superior skill but rather their superior luck. Proof: would Peter Lynch be famous if the Magellan Fund he so skillfully navigated for many years had beaten a depreciating S&P 500 by the same margin that Lynch actually beat an appreciating S&P 500? We think not.

Too Soon To Tell?

Needless to say, studies demonstrate that the performance of top quartile managers varies significantly by vintage year. For example, a domestic venture fund with a 1986 vintage that generated a 13% compound annual return would be a top performer, whereas domestic venture funds formed in the early '90s that have generated just 13% per annum are looked upon with contempt by most observers. What's that you say? You say, "venture funds formed in the early '90s are only now reaching middle age, so it's too soon to tell." Perhaps, but judging from how much money big pension funds in particular are shoving at PI firms posting strong results on early '90s-vintage funds, the eminently sane notion of gauging managers' skills by reference to vintage years may be falling into desuetude. Some institutions are clinging to this notion when evaluating PI results produced prior to the current decade but seem to be letting down their defenses when evaluating more recent returns. As for the TIFF PI Program, to paraphrase the famous ad, "We shall drink no wine before its time."