Investors must reckon with unknowns

[SINGAPORE] In a classic example of impenetrable political doublespeak, former US defence secretary Donald Rumsfeld once rambled on about “known knowns”, “known unknowns” and “unknown unknowns” when referring to the state of the world, Iraq, Osama bin Laden and terrorism.

Rather than try and figure out what he was talking about, what’s interesting is to ask whether his framework can be applied to the stock market as we head into 2026.

If you were to ask equity strategists for their views, the answer would likely follow these lines: The outlook is positive because the US economy looks okay, inflation looks to be under control, interest rates are heading down, and earnings will surprise on the upside because of artificial intelligence (AI).

It would be tempting to say these are known knowns, but the reality is that the only known known is that analysts have a vested interest in always leaning towards painting a positive picture when delivering their opinions, lest they incur the ire of their superiors or lose clients.

The truth is that the known unknowns far outnumber what is known (and investors should know this).

The era of steep rate hikes is behind us – at least as far as the early part of 2026 goes. The US Federal Reserve just cut rates again and now projects only one more cut in 2026, signalling a long plateau rather than a rapid return to near-zero rates.

What’s more, inflation has come down, but not all the way back to the pre-Covid-19 level. Forecasts see global inflation trending lower, towards roughly 3.6 per cent in 2026, with advanced economies in the mid-2 per cent range.

The first big known unknown is whether inflation’s “last mile” will behave. Wage stickiness, geopolitics and energy volatility could push inflation higher, forcing the Fed into a more cautious stance.

On growth, baseline projections point to a world that’s slowing but not necessarily stalling. However, note that JPMorgan’s latest estimate is still a 35 per cent chance of a US and global recession next year.

AI is the big, measurable spending story. Global AI capex is expected to exceed US$570 billion in 2026, driven by data centres, chips and network upgrades.

A handful of mega-caps remain the dominant beneficiaries, and earnings expectations reflect it: S&P 500 profits are projected to grow in the mid-teens, with AI-linked firms contributing a disproportionate amount.

However, what we don’t know is whether the AI game has already been priced in. The S&P 500 continues to hit record highs and trades around 24 times forward earnings – rich versus its decade average near 19 times.

Moreover, AI’s revenue payoff remains uncertain. Capex is gigantic, but the timeline for broad-based monetisation is unclear. If productivity gains disappoint, 2026 could see scepticism setting in – as it already appears to be, judging by the Nasdaq’s performance in the past few weeks.

Meanwhile, the true surprises – the unknown unknowns – could come from anywhere: a sudden regulatory clampdown on AI, a major cyberincident, a geopolitical shock, or a financial accident in private credit or shadow banking.

Another unknown unknown is if the Fed’s “put”, or explicit guarantee to bail out Wall Street in case of an epic crash, fails to have the intended impact if ever it needs to be exercised. Markets have grown so used to believing that the Fed will backstop any major collapse that if this fails, nobody knows what the end result might be.

As things stand now, investors should therefore position for the known knowns, but be aware of all unknowns – both known and unknown.