The Investment Case — Talk Track

We're overweight banks — largest sector position, about a fifth of the fund. But I want to give you the honest version of why, because the popular version of this argument is wrong and it will get picked apart.

The popular version says banks are cheap against their own history — 1.6x book against a long-run 2.1x, 11x earnings against the Nifty's 18x — and that they'll compound back to it. I'd avoid that argument. Aditya Narain, who ran research at Citi India and covered this sector for three decades, pressure-tested our work in September and called mean reversion the single biggest fallacy in the private-bank case. His point is that the means themselves have moved. HDFC traded at five to six times book when it earned 17–18% on equity in a 20% credit-growth economy. It now earns 14–15% in a 12–14% economy. Nothing about that is cyclical, so nothing about it reverts.

So here's what I'd actually say. First, the demand is real but it is catch-up, not take-off. System credit is growing around 19%, up from about 10% a year ago. But nobody borrowed for five years, so part of this is deferred borrowing coming back. Part is working capital and inventory — when steel prices rise, a dealer needs twice the financing for the same tonne, and credit rises with no real activity behind it. A steadier 16–17% for five or six years is the healthier outcome, and it's the one to underwrite. The market has not paid up for the 19% print, and I think the market is right.

Second, the funding constraint has genuinely broken. The RBI opened an FCNR(B) window in June and took the hedging cost onto its own books; banks raised about $120–130 billion in twelve weeks, roughly 4.5% of the entire deposit base. Deposit growth went from 12–13% to about 14.7% and system liquidity swung to a surplus. That's real. But it cuts both ways: the country now has a defined dollar liability sitting inside a headline reserve number that nobody nets off. Every bank is individually hedged, so no bank carries currency risk. What it does is raise the beta on the sector — in a good cycle it's forgotten, in a bad credit or currency cycle it becomes the story.

Third, the balance sheets genuinely are the cleanest in a generation. System bad loans have gone from 11.4% in FY18 to about 1.8%, with capital at record highs. That part is not contested by anyone. But note the asymmetry: asset quality can stay where it is, it cannot improve much from a twenty-year low. So the upside case for bank earnings is mostly volume at stable margins and stable credit costs. That's a decent outcome. It is not one the market pays a premium for.

So the honest framing is this: we hold banks as a diversified, liquid way to express an India catch-up trade. India has been flat for two years and has underperformed badly; a catch-up is plausible and could arrive quickly — a fall in oil toward $70 would be the obvious trigger, because it eases the currency and the inflation problem at once. Banks are the simplest and most liquid way to own that view. We'd size it for a 10–15% move. What we would not tell you is that banks are cheap against their own past and will compound back to it.

One more thing worth being straight about, because it's the part of the bear case I find hardest to dismiss. Banks were supposed to own India's platform economy — they had the customers, the accounts, the payment rails. They lost it to the brokerages, the asset managers and the wealth platforms. Those are P&L businesses with operating leverage to a market rally; a bank is a balance-sheet business and is not. If the market runs 10%, the brokerages run 30% and the banks run 12–14%. So if what you want is leverage to an Indian equity rally, the banks are not the cleanest instrument. What they are is a liquid, well-capitalised, low-drawdown way to hold the view — which matters a lot for how this fund is run.