By Jaume Argerich, Director, Aniol family office
Aniol is a family office that invests in European Search Funds since 4 years.
To make it easier for you to read this article, we suggest downloading it in PDF Format.
Every new contact between investor and searcher starts with a Private Placement Memorandum (PPM). PPMs are rather dull, as they all look remarkably similar. Many investors scroll directly to the searchers’ credentials section, which is the most differential part.
However, all PPMs contain one common ratio. That number is the 35.3% internal rate of return (IRR) from the Stanford 2022 study. This is a key selling point for the asset class. If true, it is a return that investors do not see often. As such, it is widely advertised in PPMs.
As an investor in international search funds, I searched the 11 PPMs reviewed in the last month and all contain the 35.3% percentage high up in the executive summary. That is misleading as those returns are not from the international market. The IRR for international search funds (global excluding US and Canada), from the IESE 2022 study would be a much lower 19.4%. However, this ratio appears only in half of the PPMs, mostly buried somewhere in the annexes.
What are then the returns we should expect from international search funds? The comparison of both studies reveals some interesting differences and facts enough as to take IRRs with a pinch of salt.
A first difference is the percentage of funds that fail to acquire a target. The possibility to fail is 34% in the US and Canada and 26% on the international market. The ratio by country within the international market can vary a lot, with northern Europe and Latin America showing similar rates as the US and Spain being as low as 14%. The interpretation of this is disputed. Are investors more rigorous in the US or are searchers more efficient in Spain? If we are guided by the IRR, it would look like individual failure ends up being positive for the group. The higher the percentage of funds that do not complete an acquisition, the higher the IRR of those who close one. Search capital is only 4-4.5% of the purchase price, so losing it is not a big deal when we compare it with the cost of a failed acquisition.
A second difference is timing. The US market is in a phase of consolidation. The first search fund was launched in the US in 1984, and the model took off in 1996 when they were more than 5 search funds launched in the same year. Since then, the asset class grew gradually and has been tracked biannually by Stanford.
By contrast, search funds were slow to spread to other countries. The first search fund outside of the US and Canada appeared in 1992, but 20 years of stagnation followed until the asset class started to take off in 2012. Given that the full cycle can take 8 years, the amount of exits outside of the US and Canada is very small.
In effect, the US and Canada study contains information on 136 search funds that concluded the process (58% of the sample), while as the international one contains only 14 (20% of the sample). There are too few actual observations on the international market to validate any conclusion.
Why you should not trust IRRs
It is not advisable to make sound decisions on an IRR that is made up of 80% of estimates. The US and Canada figure has a higher share of real cases but in both markets we should take out the impact of estimates for search funds that are still in operation. Unfortunately we do not have IRR for operating and completed funds but we do have the return on investment (ROI). Operational search funds are valued according to its operator exit estimates. If we distinguish actual from estimates, we find out the following curves (see figure 1).
Figure 1. Acquisition outcomes in returns on investment for the US and Canada and the international market, differentiating Op (funds still in operation) and Ex (completed funds that have exited)
International (in red) contains 14 completed funds (continuous line) and 55 operating funds (discontinuous line). US and Canada (in blue) contain 136 completed funds (continuous line) and 97 operating funds (discontinuous line).
The curves for operating funds distort the picture. Operating funds estimated returns follow a normal distribution clustered around the 1-2x ROI segment. Probably, many of those operating funds are in their early stage and with time they will become either failures or successes. Remarkably, none of the 55 international funds believes it will close with a loss. In reality, that was the outcome for 30% of the 136 completed funds in the US and Canada. The problem with those estimates is that since they make 80% of the international study, they hide the reality of actual returns.
Focusing on completed cases, the difference between US and Canada and the rest of the world is not so big. The curves follow a pattern with higher proportion of failures, a minority of returns in the 1-2x ROI segment (as low as 8% of the total in the US and Canada) and many cases above 2x ROI. It would look like the US and Canada have a higher risk higher return profile than the rest of the world.
Comparison with venture capital
In any case, actual returns for search funds have a high share of failures (out of 100 search funds raised, 32 do not complete acquisitions and 20 report losses), many cases of stellar returns (23 make more than 5x ROI) and some decent ones (20 between 1 and 5 times ROI). Of course, the impact of the ones that do not close an acquisition is much smaller in euros.
Many of the current investors in search funds come from venture capital and will find that distribution of returns familiar. Are search funds returns going to mirror those of venture capital? Venture capital returns follow a power law distribution, with most losses and a long tail to the right with a very small minority of extraordinary returns.
In figure 2 there is a comparison between the returns of the 150 search funds completed worldwide and venture capital, with a sample of 35 thousand US deals of the last decade. The search funds curve has 19% less of losses and 17% more successes in the 5-10x ROI area. Both asset classes have a similar percentage of super success cases above 10x ROI (8% in search funds to 7% in venture capital) but probably the tail of venture capital is longer. The subsegment of >10x ROI returns in venture capital contains half of the cases between 10x and 20x and the other half above 20x. There is no data for search funds but probably they have a lower share of spectacular returns.
Figure 2. Search funds and venture capital returns.
Conversely, total failures should be lower with search funds. Actual search funds report that the 29% of failures contain 16% of total losses and 13% of partial losses. Partial data for venture capital suggest that 75% of the returns below 1x are total losses.
What investors and searchers can learn
According to a McKinsey study on venture capital during the last 20 years, long term IRR for the venture capital was 15.7%, with high seasonality across economic cycles (2022 and 2023 both reported losses).
Search fund investors should expect the international market to move gradually towards the US and Canada market. The 35.3% IRR has 40% of estimates in it but the study mentions that actual returns are higher than estimated ones. Full convergence may never happen, as the international market appears to be more prudent. We can see that in the international market there are fewer extreme failures and successes with the same average size of deal and company, probably due to the lower share of software deals.
For international search fund investors, full or partial convergence with the US is not a big deal. If the market stays on the current 19.6% IRR, search funds returns would still be better than venture capital with less dispersion between individual investment and less cyclicality.
However, international searchers should consider the IRR implications with care. The current IRRs, with all their limitations, define an area between 20% and 35% which fits exactly the vesting of their third tranche of equity. If the international market gets to US and Canada level, they will on average vest all their shares. If not, they would stay with the first two tranches.
Many investors worry that the sector may follow the path of venture capital. Venture capital suffered an excess of capital driving up valuations and leading to a boom-and-bust cycle. It is difficult to see a similar bubble in search funds. The underlying asset is an SME, it is therefore unlikely we will see anyone betting on insane valuations based on unlimited scalability. Furthermore, there is ample supply of SMEs and succession problems to absorb additional investor money without pumping up valuations. Finally, the rise in interest rates of the last years did not have the killer impact seen in venture or in institutional buyouts.
The gap in target IRR between searchers and investors should rather be a cause of concern. Investors would be happy with 20% compared to other asset classes, but searchers are targeting 35% to vest the third tranche taking extra risk with the investors’ capital. Maybe this difference could create a misalignment between searchers and investors. It may force searchers to overlook companies that have good potential but not as high as to deliver 35% returns.
The 2024 update for both studies is due in a few weeks. We will then be able to see whether convergence is on the way.
