????長期以來，風投行業一直以“基金”為單位來衡量投資表現；風投公司圍繞特定策略、募集成立相互獨立的資本池——基金，然后遵循這一策略，對公司進行投資。“基金”概念在風投行業歷來非常重要，因為投資回報正是基于基金計算的。而像美國康橋咨詢公司（Cambridge Associates）這樣的研究公司，每天的工作就是比較不同基金的相對表現。

????但是，風投行業以基金為單位的考評體系賴以存在的根本假設正在慢慢失效，即人們可以、且僅可以投資于某一只基金。

????我來具體解釋一下。現在，風險投資業內出現了兩個并行不悖的趨勢，導致以基金為單位的考量體系正在慢慢失效。第一個趨勢是，有限合伙人（LP）越來越多地開始圍繞品牌風投公司整合自己的投資資產。越來越多的資金流向了更少的風投公司。在該趨勢出現之前已經形成并與其并行發展的是，品牌風投公司紛紛創建可供選擇的不同類型基金，運用不同的策略進行資本配置。

????十年前，風投公司每次只通過一只基金投資。可如今，大型風投公司基本上都有自己的種子基金、早期基金、成長基金、中國基金、印度基金和專門基金（如移動、數據）等等。

????關鍵在于，有些風投公司不允許有限合伙人選擇他們最感興趣的基金并只投資這一只基金。他們強制進行更廣泛的資產配置，要求有限合伙人投資多只基金。

????業界存在兩種做法。有時，一般合伙人會拿出最有吸引力的基金（所有投資者都想投資的基金），規定有限合伙人如果想投資這只基金，就必須同時投資他們旗下的其他基金。或者，一般合伙人會說，有限合伙人投資的不是某一只“基金”，而是“特許權”。它意味著有限合伙人把資金交給了風投公司，由風投公司決定如何將這筆資金分配到不同的基金中。不同的方法，同樣的效果。

????所有這些意味著什么？簡單來說，就是基金層面的投資回報并不總是那么能說明問題，因為有限合伙人并沒有機會單獨投資這只基金。如果有限合伙人只能通過投資一籃子基金才能獲得和這家公司共同投資的機會，那么真正能反映投資回報水平的，只能是一籃子基金的投資回報數據，而不是單一的某一只基金。但由于很難知道這些籃子的具體構成，以單只基金為單位的評價體系仍在沿用，只是對一批業內規模最大、知名度最高的品牌基金來說，這個評價體系的重要性正在日漸降低。這個過程或許是緩慢的，但它預示著整個行業更普遍的趨勢。

????本文作者拉里?陳 (@larryvc)是波士頓風險投資公司Volition Capital的合伙人。他的博客是Thinking About Thinking。

????The unit of measurement in the venture capital industry has long been the "fund" -- a discrete pool of capital that a firm raises around a particular strategy, and then deploys by investing into companies aligned with that strategy. The fund has had longstanding significance in the venture industry because it's how returns are calculated. And research firms like Cambridge Associates have been built around comparing the relative performance of various funds.

????But the utility of the fund as the metric of measurement for the venture capital industry relies on a fundamental assumption which is slowly, but increasingly untrue: That you can invest in that fund, and only that fund.

????Let me elaborate. There have been two parallel trends in the venture capital industry that are eroding the utility of the fund as the unit of measurement in the industry. The first trend is that LPs increasingly are consolidating their positions around the branded venture capital firms. More dollars are flowing to fewer firms. Prior to, and in parallel to, that trend, the branded venture firms are creating a menu of different types of funds with distinct strategies with which to deploy that capital.

????Ten years ago, venture capital firms were investing out of one fund at a time. Now it is not uncommon for the largest firms to have their seed fund, their early stage fund, their growth fund, their China fund, their India fund, their specialty fund [mobile, data, etc.] and on and on.

????Here is the key point: Some of these firms will not allow LPs to select the fund that most interests them and invest in that fund on a discrete basis. They force a broader asset allocation into multiple fund vehicles in order to invest with that firm.

????The logic works in one of two ways. Sometimes the GP will hold out the most attractive fund – the fund that every investor wants to get into – and stipulate that in order to get into that fund, the LP must invest in their other funds as well. Other times, the GP will say that the LP isn't investing into a "fund", rather they are investing in a "franchise." This means they allocate capital to the firm, and the firm may decide the relative allocations to the various funds. It's another approach to the same end result.

????How does this all net out? It simply means that fund-level returns aren't always relevant because LPs never had the opportunity to invest discretely in that fund discretely. If LPs had to invest in a basket of funds to invest with that firm, then the true unit of measurement of returns is the basket, not an individual fund. But, it's impossible to know how these baskets take shape so the individual fund persists as the unit of measurement, despite its increasing irrelevance with a number of the largest and most branded funds in the industry which may slowly but surely portend a broader trend for the industry as well.

????Larry Cheng (@larryvc) is a partner with Boston-based VC firm Volition Capital. He blogs at Thinking About Thinking.