There were $13.1 trillion in Assets Under Management (AUM) in the Alternative Investment Industry as of June 30, 2023. The AUM have grown at nearly 20 percent per annum since 2018 (McKinsey Report 2024). The split of these assets among strategies is approximately $3.6T in Buyout, $2.8T in Venture Capital, $1.3T Growth Equity, $1.5T Private Debt, and $3T Real Estate & Infrastructure (McKinsey Report 2023).
The annual budget of these firms is usually funded with a yearly management fee of approx. 2% p.a. of the AUM. The allocation of this budget varies between the firms but, on average, approx. 65% of the budget is dedicated to personnel costs ($170B or 1.3% p.a. of the AUM). This strongly depends on how much of the management fee is used to fund the GP commitments, which tend to be around 2-3% of the fund size (one time, not annual).
According to my estimation institutional investors spend approximately $13B per annum on technology (0.1% of their AUM or 5% of their budget). This amount has grown quickly over time. When I started in the VC industry 12 years ago many firms used spreadsheets as a CRM (to be fair, the firm where I started my career was much better equipped than their smaller peers). Now virtually everyone has a CRM and in many cases a full stack of applications running on top of it.
So the question then becomes, what is the right way to estimate the TAM of AI in the institutional investment industry? Is it the current technology budgets or the future ones? Or is it the salary budgets? I believe the answer is somewhere between the technology spend of $13B (lower bound) and the workforce spend of $170B (upper bound). The estimate is not very precise but, as I heard one GP say once: It’s bloody big enough.
In my view, AI is going to affect everyone without exceptions. The only question is how long it will take. When I hear some investors say that AI will not affect their stage or their type of strategy I think about all the companies that succumbed to technological revolutions in the past. These types of statements have historically not aged well. Let’s see what history teaches us:
Since 1900 until today the agricultural machinery has taken over the agriculture industry. In the US, this has generated a 4x increase in farm yields while reducing the amount of workers employed by about 75% (Jayson Lusk). In another study it has been estimated that the agricultural output per worker grew 16 times between 1948 and 2017 (USDA). A similar thing has happened in the production industry (Mark J. Perry). The U.S. increased industrial production by 4.3 times from 1947 to 2011 while decreasing the number of employed workers by 30% leading to approx. 6x increase in output per worker. Robotics has been taking over discrete manufacturing step by step and AI is going to speed up the process. I believe that AI will produce a similar outcome in all white collar industries too.
This is where speculation kicks in but if the reader has scrolled so far down I might as well give a few ideas.
Guillem Sague
CEO of CarriedAI