Infrastructure and PSU funds: riding the government capex cycle

These are the funds that turn a newspaper headline into a portfolio position. Capex is rising, the government is building, the PSUs are being reformed — and the fund that owns exactly that has just returned 40%. Both halves of that sentence are usually true. The second is the reason to be careful.

What you are buying

Infrastructure funds are thematic schemes that hold the construction chain: capital goods, cement, engineering and construction, power, roads, ports, railways, defence manufacturing, and the banks that finance them.

PSU funds hold state-owned enterprises — public-sector banks, oil marketing companies, power utilities, defence undertakings, mining and metals. Some schemes restrict themselves to PSU equity; others hold PSU debt, which is an entirely different instrument with entirely different risk.

The two overlap heavily, because much of India's infrastructure is state-owned. Owning both is usually owning one thing twice.

As thematic schemes they must hold at least 80% in the stated theme, and SEBI's revised categorisation adds a constraint worth knowing: a sectoral or thematic scheme's portfolio overlap with the AMC's other equity schemes (large-cap funds excepted) is capped at 50%, measured on portfolio values — a direct attack on the practice of launching a "theme" that was substantially an existing fund.

The cycle you are actually betting on

Infrastructure is the most cyclical part of Indian equity, and the cycle is long, policy-driven and violent at both ends.

The pattern has repeated: a capex upturn begins, order books swell, earnings surprise upward, and the sector runs for several years. Valuations re-rate from depressed to demanding. Then the cycle turns — interest rates, a fiscal consolidation, a slower award pipeline — and the sector spends years going nowhere while the broader market compounds without it.

The 2003–2008 infrastructure boom is the reference case, and so is what followed it: a decade in which infrastructure investors made very little while diversified equity did well.

PSUs add a second variable: policy. A PSU's earnings can be changed by an administrative decision — a fuel subsidy, a dividend directive, a divestment plan, a pricing formula. That cuts both ways, and it is not risk a fund manager can analyse away. It is also why these stocks trade at a persistent discount to comparable private businesses: minority shareholders are not always the first priority.

The valuation trap in these funds

Here is the mechanism that catches people, and it is worth understanding once properly.

Cyclical companies look cheapest at the top of the cycle and most expensive at the bottom. At the peak, earnings are at a maximum, so the P/E looks low. At the trough, earnings have collapsed, so the P/E looks enormous or meaningless.

So the naive signal — "this sector is on 12× earnings, it is cheap" — reliably points the wrong way in cyclicals. The P/E on an infrastructure fund's factsheet after three strong years is not a discount; it is peak earnings capitalised.

Combine that with the launch cycle — new thematic schemes appear after the theme has performed, not before — and you get the standard outcome: retail money arrives at maximum earnings and minimum apparent valuation.

If you want the exposure

• Check what you already own first. Diversified funds and the broad index already hold large-cap capital goods, power and PSU banks. Adding a thematic fund does not diversify; it doubles. Measure the overlap.
• Size it as a satellite you can watch halve. 5–10% of the equity allocation is the commonly cited ceiling, and it is a sensible one.
• Have a thesis with a time horizon attached. "Government capex is rising" is not a thesis unless you can say for how long and what would falsify it.
• Decide the exit in advance. The narrative will still sound compelling when the cycle has turned — that is what makes cyclicals so hard to sell.
• Rebalance out of it mechanically. If the sleeve doubles, trim it back to target. This is one of the few places where a rebalancing rule reliably harvests a cycle rather than riding it down.

Pitfalls to avoid

• Buying after the run. The returns that made you notice the fund are the reason for caution, not the reason to invest.
• Reading a low P/E as cheap. In cyclicals it usually signals peak earnings.
• Owning an infra fund and a PSU fund. Substantially the same bet.
• Confusing a national priority with an investment. India will build the infrastructure. That does not mean today's price for the builders is a good one — the two questions are unrelated.
• Running a SIP into it and calling it discipline. Averaging into a concentrated cyclical does not diversify it, and the benefit depends on a recovery whose timing is exactly the thing you cannot know.
• Forgetting the exit tax. These are equity-oriented, so trimming a position that worked triggers capital gains — the hesitation that follows is how gains get given back.

Key takeaway

Infrastructure and PSU funds are a leveraged bet on the capex and policy cycle, and the cycle is long enough that being early and being wrong feel identical for years. The specific trap is valuation: cyclicals look cheapest exactly when earnings have peaked, which is also when the schemes get launched and marketed. Own the theme only as a small satellite on top of a diversified core, with a falsifiable thesis and a rebalancing rule that trims it after a run — and check first how much of it your existing funds already hold.