Domaz

I believe that capital markets are changing in a way that is favorable to those of us who are inspired by the art and practice of capital allocation. The number and complexity of financial instruments is exploding and, with them, the number of opportunities for thoughtful investors to build and express edge. 'Edge' is the defining property of the Investment Manager of the Future where 'Fundraising' has been the defining property property of the Investment Manager of the Past. The Investment Manager of the Past ('IMP') is adapted to cumbersome fundraising, extensive diligence, few assets, and economies of scale. You needed a large office in midtown to raise a >$1B fund, and you needed a >$1B fund to hire fifteen analysts and associates out of Wharton. Deals were 20% investment analysis and 80% execution so you needed an army of lawyers, accountants, and back office staff to execute on any reasonable strategy. Then you needed management, bureaucracy and process to manage the organization. The more you raised, the more you could raise from the larger endowments and institutions you met at Goldman conferences and expensive investor days. The more you raised, the more diversification and AUM mattered to the fund's leadership. The more you raised, the more you were a beta instrument fighting for a relatively thin spread above median. The economies of scale were inescapable. This is, of course, a caricature. There are many extremely high quality alpha shops in the world that do not operate this way but I would wager that more than 60% of AUM lives in these types of entities. The Investment Managers of the Future ('IMF') will adapt to Internet Capital Markets and LLMs. They will fundraise directly on Internet Capital Markets, where track record and quality speak for themselves. The opportunity set will not be limited to a few thousand stocks or a few hundred banker-run sales processes each year but the hundreds of thousands of assets (and maybe even millions of assets) that trade on Internet Capital Markets. They will not need to hire fifteen analysts from Wharton because ChatGPT / Cursor can put 10,000 analysts in your pocket. The incentive will no longer be size and prestige but pure alpha relative to a benchmark and so these businesses will double down on the only thing that has ever lead to alpha: building and expressing edge. The economies of edge become inescapable as you are forced to compete for global, permissionless capital on the Internet. These funds will be smaller operations focused on special situations, niche markets, and specialized knowledge. The two structural forces driving this change are Internet Capital Markets and LLMs. Internet Capital Markets mean that every asset in the world is going onchain. There will be funds in NYC that specialize in underwriting Egyptian Auto Loans, or buying Argentinian farmland from laptops in Lisbon. There will be funds that build algorithms to trade music royalties and LLMs to underwrite structured mortgages around the world. We are even creating new asset types through new financial primitives like decision markets and governance markets. Every election and key corporate governance decision will lead to the creation of new assets that focused funds can study and trade. Internet Capital Markets also disrupt one of the core barriers to entry in Investment Management: fundraising. I believe that we are only a few years from a world where investment managers can compete for capital on the Internet on platforms where they can show proven track records to select LPs and bake legal structure directly into smart contract code. The main barrier here is investor protections — LPs need to be able to trust that there is sufficient investor protection infrastructure for them to participate, and we are making significant headway on this product as an industry. LLMs like ChatGPT / Cursor allow small teams in 2026 to produce multiples of the analysis that large teams could create in 2018. You don't need to spend 100 hours building a model in Excel because Cursor can build you the same model in less than an hour using Python. You don't need to spend 2 weeks learning from industry experts if you can focus through a 5 hour active Q&A session with ChatGPT (and then you can spend 2 days with experts on the finishing touches). An entire world of powerful modeling techniques that were only possible at large institutions with armies of analysts are now at your fingertips (you can monte carlo anything these days, and I recommend that you do). 5 years ago, I would have preferred 100 college grads who needed to be managed to 5 high agency teammates who could manage themselves. There were just too many presentations, spreadsheets, and emails; there was too much busywork for a small team to get through. Today, competent people with high agency can get weeks of work done in hours and have the judgement and experience to add value by themselves. In 2026, I would pick the 5 high agency teammates over the 100 recent college grads any day. LLMs have unleashed a supermassive gravitational pull towards lean, efficient firms. This is an easy prediction of Coase's Theorem, which predicts that firm size will depend on the relative trade-off between the costs of operating a large business (hierarchy, principal-agent problem, motivation) versus the benefits of operating a larger company (economies of scale, processing power, dozens of departments with specific knowledge). LLMs plus Internet Capital Markets reduce the economies of scale for Investment Managers and tip the scales towards smaller alpha shops. I believe this is the investment management industry returning to its original mandate of making predictions about the future and allocating capital where it is most productive. John Maynard Keynes put it best when he said that 'The social object of skilled investment should be to defeat the dark forces of time and ignorance which envelop our future.' Capital allocation matters because resource allocation follows market prices and company building follows fundraising. Investors are supposed to do the hard work of understanding the world and attempting to pierce the veil of the future even if only slightly to develop and express edge, defined as predictions and opinions that are both correct and different from consensus. This edge then shows up in capital markets (through your investments) and in the material world as the economy decides how to allocate its limited resources — engineers, laborers, scientists, buildings, commodities, machinery — by looking to capital markets. I expect a global ecosystem of alpha shops using LLMs to build edge will eventually do a better job at producing edge than a correlated group of 'economies of scale' / beta-tracking firms all located within fifteen blocks of each other in midtown Manhattan. This essay is a natural corollary of our thinking around the Internet Financial System. You can give our initial essay a read below if you are interested: