A Mathematical Analysis of Value-Creation Attribution in Search Fund Projects
Daniel Lazier, Jacob Thomas, A. J. Wasserstein. (2026)
Yale Case, no number printed, January 12, 2026
Sample. "we contacted a broad range of serial ETA investors with a request for unmasked data. Ultimately, we received information from six investors about companies in their portfolios that have fully exited; there are no interim mark-to-market valuations, and all of the firms are classified as traditionally funded searches." (p2). "we created a comprehensive dataset of 59 unique observations" (p2), de-duplicated across investors; 48 remain after dropping exits above 19.9x EBITDA. "our dataset includes some observations from outside the U.S. and Canada" (p2). Time window: not stated; mean holding period 4.8 years (figure). Included and excluded. Fully exited, traditionally funded only. Excludes unrealized companies, self-funded deals, companies with negative EBITDA at exit, and leverage ("We exclude leverage from our analysis intentionally", p1). Organic and acquired EBITDA growth are not separated. Enterprise value only; equity returns are not analysed. Headline numbers. "EBITDA increases from roughly $158 million at entry to $254 million at exit, and the multiple rises from 6.3x to 15.6x." (p3). "approximately 20% ($605 million) to growth in EBITDA, and the remaining 80% ($2.4 billion) to multiple expansion." (p4). "Revenue increases from approximately $634 million at entry to $1.4 billion at exit, and the EBITDA margin contracts from 25% to 19%." (p5). "The total EV creation across our sample can be viewed as originating from three sources: 80% from EBITDA multiple expansion, 39% from revenue growth, and -18% from EBITDA margin contraction." (p6). "only 30% of the sample see EBITDA margin expansion, while 70% experience a reduction in EBITDA margins." (p8). "The EBITDA multiple decreases for only eight of the 59 firms" (p7). Coverage claim. "according to the 2024 Stanford Graduate School of Business Search Fund Study, 168 traditional search funds have exited. In other words, we are capturing 35% of the known ETA population that has exited." with the caveat "we are not making a direct apples-to-apples comparison." (p2). Limits. Authors name omitted variable bias, sample selection, survivor bias ("this group mainly consists of successful companies") and limited observations (p17 to p18), and note exits at 20x or more "probably relate to metrics other than EBITDA". Reader's notes: the 80/20 headline is a dollar-weighted aggregate that includes 11 exits at 20x or more; excluding them the split is 64% / 36% (figure, p3). The attribution is sequential (entry multiple applied to EBITDA change, exit EBITDA applied to multiple change) and a different ordering would move the split; the authors do not discuss this. One multiple regression on the 48-observation subsample (R squared 0.476, figure). Related Yale Insights summary is listed in section 2.