Ask a general partner where returns come from and you will hear about sourcing, judgement, pricing and the patience to hold. All true, and all concentrated in a handful of decisions per year. What rarely comes up is the machinery that runs continuously underneath those decisions: the drawdown that goes out, the wire that gets reconciled, the valuation that gets marked, the statement that reaches an investor. Nobody puts that machinery in a pitch deck. It still shows up in the numbers.
Operational alpha is the return you keep — or lose — for reasons unrelated to investment skill. It is unglamorous and it compounds, which is an unusual and underrated combination.
Where operational alpha actually hides
It is tempting to treat operations as a cost line to be minimised. That framing misses the point, because the effect is not mainly about cost. Four mechanisms matter more.
1. Idle capital
Between the moment a fund decides it needs capital and the moment that capital is usable sits a sequence of manual steps: computing each investor's share, drafting notices, sending them, chasing the ones that bounce, and matching money as it arrives. Every day in that sequence is a day of capital not deployed, or a day of drawn capital sitting in a current account dragging on IRR. The arithmetic of an internal rate of return is unforgiving about timing in a way that it is not about small differences in exit price.
2. Marks that lag reality
If a portfolio company's numbers reach you six weeks after quarter end, and the mark is set two weeks after that, then decisions about reserves, follow-ons and exit timing are being made against a picture of the past. The cost is not a reporting failure. It is a series of slightly mistimed decisions, each defensible in isolation.
3. The close as a bottleneck
When the quarterly close is a reconstruction rather than a review — waterfall rebuilt in a spreadsheet, capital accounts recomputed, statements retyped — the most experienced people in the finance team spend their quarter on arithmetic that a computer settled decades ago. That is the real cost: not the hours, but whose hours.
4. What the investor experiences
An investor's judgement of a manager is formed substantially through operational contact. The onboarding that took three weeks. The statement that arrived late, or arrived and disagreed with the last one. The question that took four days to answer because someone had to open six files. None of this is investment performance, and all of it shapes the re-up conversation for the next fund.
Operations do not decide whether you picked the right company. They decide how much of that decision survives the trip to the investor.
Why spreadsheets persist, and why that is not stupidity
It is easy to be superior about spreadsheets. Do not be. A spreadsheet is the most flexible modelling tool ever built, it requires no procurement, and it does exactly what the person in front of it wants. For a first fund with a dozen investors and one currency, it is genuinely the right answer.
The problem is not the spreadsheet. It is what happens as the fund family grows. A second vehicle arrives with different fee terms. An investor commits in a second currency. A co-investment vehicle appears alongside the main fund. A secondary transfer changes the register mid-year. Each of these is handled by adding a tab, and each tab is a private agreement between one person and a file. The knowledge stops being institutional. Eventually the fund is running on a system whose documentation is a colleague's memory.
The switching cost is also real, and pretending otherwise is how technology gets oversold to fund managers. Migrating a fund's history means reconciling to a number an auditor has already signed. Any platform that treats that as trivial has not done it. The way we handle it is to run one close both ways, in parallel, until the two agree to the rupee or the cent — and to treat that reconciliation as our work, not the client's.
What good operations look like
A useful test: how much of the following is true on an ordinary Tuesday, with nobody preparing for anything?
- The net asset value on screen is today's, not last quarter's.
- A capital call can be built, computed per investor, issued and reconciled without a spreadsheet leaving the building.
- The waterfall is a computation the system performs, not a file someone owns.
- Any investor's capital account can be produced on request, and it agrees with the last statement they received.
- You can say who changed a valuation, when, and what it was before.
- A portfolio company's update arrives in a structured form, not as prose in an email.
None of these require artificial intelligence, and it is worth being clear about that. They require one record instead of several, and computations that run when events land rather than when a quarter ends. AI is genuinely useful on top of that foundation — an agent can chase a missing declaration or draft a report — but an agent pointed at four disagreeing spreadsheets inherits the disagreement. That ordering matters, and it is the part most often got backwards.
For Indian AIF managers specifically: the reporting calendar and the investor-documentation burden make the gap between good and poor operations wider than in some other jurisdictions, because the same fixed process runs more often.
How to know whether you are getting better
Operational alpha resists a single headline number, which is why it gets ignored. It does not resist measurement. A manager who wants to improve can track a small set of internal metrics over time, none of which require a consultant:
- Days from onboarding start to an investor being able to fund.
- Days from capital call issue to the last rupee or dollar reconciled.
- Days from quarter end to LP statements dispatched.
- Proportion of portfolio companies reporting in a structured format on time.
- Number of restatements or corrected statements issued per year.
- Median time to answer an ad-hoc investor question.
Track those for four quarters and the picture stops being a matter of opinion. Most managers who do this discover the constraint is not effort — their teams work hard — but that the work is shaped by tools that force a rebuild every cycle.
The argument, briefly
Investment skill is scarce, hard to acquire and hard to prove. Operational competence is none of those things: it is available, it is buildable, and it is visible to the people writing the cheques. Treating it as back-office hygiene rather than a source of return is a choice — and increasingly a conspicuous one, because investors have started asking operational questions in diligence that they used not to ask.
The managers who take this seriously will not describe it as alpha. They will describe it as not losing things. That is the same thing.