Turkey's Forecasts Got More Honest, and Someone Is Paying for It

The strange thing about Turkey's vice president defending his economic forecasts this week is that he did it at the moment he was quietly admitting the last ones were wrong.

Cevdet Yilmaz on Sunday unveiled the government's new 2027-2029 medium-term program, and the headline number was a concession dressed as a forecast. He cut this year's growth projection to 3.3% from 3.8%. He let the year-end inflation forecast drift up to 28.4% — from the earlier, more hopeful path that in 2024 had promised Turkey down to single digits by this year. Then he said the forecasts were realistic. "We think our forecasts are realistic" is basically the sales pitch of anyone presenting a plan that has failed on schedule, but in Turkey the line has a specific job to do.

In most rich countries, an inflation forecast is a piece of internal budgeting: the central bank publishes it, markets shrug, and the number drifts. In Turkey, the forecast is the collateral that keeps the whole machine running. The lira carry trade — borrow cheap somewhere, park the money in Turkish lira yielding 37%, pocket the difference — only survives while foreigners believe the disinflation program will hold. The official forecast is how the state signals that belief to them. A forecast that was visibly absurd would pull the plug on the very capital the country needs to keep the program alive. So one way to read Yilmaz's "realistic" is as honesty-as-currency: the state now has a financial incentive to publish numbers that bear some resemblance to the truth, because the truthfulness is what the carry trade is buying.

The numbers themselves show why that balance is delicate. Inflation peaked at 75.5% in May 2024 and has been grinding down to 31.5% as of August. The new program sees 28.4% by the end of this year, then 21% in 2027, 13.5% in 2028, and 9% in 2029. Notably, the government now roughly matches the central bank, which in August lifted its own 2026 year-end forecast to 28% from 26%. That convergence matters: when the government's and the central bank's numbers used to disagree, it was a sign that politics and policy were pulling in different directions. Here they are admitting the same thing at the same time.

What forced the retreat was, officially, a war. Yilmaz attributes the upward revision to the Middle East conflict, which the central bank estimates added about seven percentage points to inflation through energy, food, and shipping costs, on top of a real-economy slowdown. The growth side of the ledger did the rest of the work: the economy grew just 2.3% year-on-year in the second quarter, below the roughly 2.9% projected, the fourth consecutive quarter of deceleration. Tight policy is doing its job — that is almost the point — and the job is bruising.

Which is where the investor decision, such as it is, lives. Turkey's high nominal yields have made lira assets one of the world's better carry trades this year, and foreign investors have stayed in the trade through a year of steady rate cuts. The policy rate sits at 37%, unchanged since January, against 31.5% inflation, so a lira saver is being paid a real cushion of several points — but only if the currency does not depreciate faster than the yield covers. The whole trade is a race between the central bank cutting rates toward the disinflation path and the lira sliding. Markets are pricing roughly 200 basis points of cuts by year-end; foreign managers who own the trade warn that cutting too early, before inflation is durably lower, is the way the cushion gets wiped out.

So the forecasts you should actually watch are not Yilmaz's medium-term ones. They are the central bank's interim targets — 24% for end-2026, 15% for end-2027 — and whether it defends them as it keeps easing. That is the number that sets the real yield a lira lender is being paid. Every time the government revises its plan to look more realistic, it is really booking a smaller cushion into the carry trade that is financing the program. "Our forecasts are realistic" means: the number we promised last year is not the number you should have priced, and the price of that honesty is coming out of the yield someone was counting on.