聚焦2025年主題—財政挑戰、貨幣政策、生產力與波動性 (只提供英文版)

As 2025 gets underway, I want to highlight four key themes that I believe will shape economic and financial market developments over the next 12 months and beyond. These themes in part reiterate points that I stressed over a year ago—the fact that they have played out as I expected over the past 12 months bolsters my conviction that they will be even more relevant as we move forward into this new year.

1. My first theme for 2025 is once again a reality check on fiscal policy, and its implications for inflation and bond yields. Since 2007, the eve of the global financial crisis (GFC), government expenditures and public debt levels have risen substantially across Western economies. Debt-to-GDP ratios have doubled in the United States, more than doubled in the United Kingdom, and now surpass 100% in all G7 countries except Germany. In Europe, this heavy debt burden is beginning to severely constrain governments’ ambitions, including key priorities like the energy transition. In the United States, last year the deficit was close to US$2 trillion, and President Trump’s election promises would send it even higher. It seems most likely that those promises will need to be scaled back, but even so it will be difficult to reduce the deficit significantly without some very hard decisions on Social Security and health care.

The incoming administration’s aim to improve government efficiency is laudable, but non-defense discretionary expenditures are already too small a share of the budget—they amount to less than US$1 trillion. High interest rates and a large debt stock imply that interest spending weighs heavier on the budget than just a few years ago (interest spending has doubled to 3% of GDP). Action on entitlements will sooner or later be necessary. For 2025, though, we’re most likely looking at a still very large deficit, and a fiscal policy so loose in an economy that’s already growing fast will keep inflationary pressures alive—I believe it’s a key reason why progress on disinflation has stalled over the last several months. Financial markets appear to be coming to terms with the fact that reducing deficits in the coming years is going to be very hard—which implies more upward pressure on government bond yields, coming both from stubborn inflation and from large issuance needs. 

2. My second theme is a reality check on monetary policy normalization. At the start of last year, I argued the Federal Reserve (Fed) easing cycle would be short and shallow, with at most 125-150 basis points in easing. I now believe it might be even less. The Fed last month signaled it expects just two more rate cuts this year, but I feel this easing cycle increasingly looks like it might be set for an extended pause if not an outright end. The just-released December jobs report showed that the labor market remains strong. Nonfarm payrolls again beat expectations at 256,000; the unemployment rate ticked down to 4.1% with unchanged participation; and wage growth momentum remains robust, at about 4% over the past three and six months. We’ll see how fiscal policy and inflation developments shape up, but if we keep seeing these kinds of numbers, the Fed will be in an increasingly difficult position.

This also bolsters my long-held conviction that neither policy interest rates nor bond yields are going back to the exceptionally low levels that we’ve experienced between the GFC and the pandemic. Those very low rates were not the new normal but a historical anomaly, and we should be looking instead at the decades preceding the GFC. This would imply that the neutral policy rate—as I have argued before—is probably around 4% and that 10-year US Treasury yields (UST), in normal circumstances, would be at least 5%.

3. My third theme for the year is productivity growth and sorting the substance from the hype in technology innovation, particularly in artificial intelligence (AI). The hype on generative AI (GenAI) has run so far ahead of reality that a resetting of expectations now seems inevitable. It’s going to take a lot more progress before GenAI can move the needle on productivity at the macro level—progress in addressing the hallucinations problem will be especially important.

Meanwhile, though, it does look like companies’ efforts to adopt prior waves of digital innovation and AI, including machine learning, and robotics, are beginning to bear fruit. US productivity growth averaged 2.5% over the five quarters ending in the third quarter 2024, compared to 1.5% over the previous ten years. Faster productivity might help put the US economy on a stronger growth path going forward—and that in turn implies higher real interest rates.