<iframe src="//www.googletagmanager.com/ns.html?id=GTM-PFR3SFR" height="0" width="0" style="display:none;visibility:hidden">

The Federal Reserve’s imperfect toolbox

Mar 13, 2026 | Atul Bhatia, CFA


Share

The U.S. central bank wields a powerful hammer with its ability to move interest rates. Despite its strength, though, we think the central bank is poorly suited to address key concerns arising from high oil prices and the rollout of AI.

Let's start the conversation

If you'd like to discuss anything in more detail, please reach out here:
Contact Us

Home improvement

By moving interest rates, the U.S. central bank can help drive the cost of credit, thereby spurring or discouraging investments, and it can push asset prices higher by engaging in longer-term bond purchases. These are undoubtedly powerful economic tools.

But the Fed is not omnipotent. And as we survey the economic landscape, most of the key issues we see are, if not immune, certainly highly resistant to Fed intervention. The result is that while guessing the direction of Fed policy will be critical for certain types of cash management decisions, in the larger investment context, a few 25 basis points moves here or there may well be overshadowed by other policy and macroeconomic factors.

As we see it, the biggest tasks for U.S. and global policymakers are responding to a potential oil supply shock and dealing with the implications of AI on workers—both as economic units and as members of society. Neither of these factors, however, is amenable to correction by the Fed.

Stagflation nation

Oil price shocks represent a risk of stagflation, a period of above-target inflation coupled with slow, or even negative, economic growth. This is both theoretically and practically tough for a central bank. Inflation calls for higher rates, contractionary environments call for lower rates, and policymakers risk exacerbating a negative trend with any move.

The real problem, in our view, isn’t that the Fed is forced to pick a side between contraction and inflation. The real problem is that the Fed’s policy levers are largely useless to deal with either condition.

Let’s start with the contractionary effects of an oil supply shock. A contraction generally calls for looser policy and easier credit to spur investment and hiring.

But if energy can’t be had at an economically viable price to produce or transport goods, what does it matter if funding costs are a few basis points—or even percentage points—lower? A logistics firm isn’t going to go on a truck-buying spree if high diesel prices make it impossible to run the machinery profitably. Even firms that have access to energy supply at a reasonable price aren’t immune to an energy shock; there’s no point in them building a new factory if their customer base is struggling to pay for gas. Rate policy works well to encourage growth when cost of capital is the binding constraint on investment, and that’s just not the case today, in our view.

Dealing with supply-side inflation is equally problematic for the central bank. Rate hikes are the usual policy tool used to deal with rising prices, and the idea is to restrict credit and growth to bring supply and demand into better balance. But when there is a restriction in something as fundamental as energy, the amount of contraction required to bring balance would be crushing for the economy and the country, in our opinion. It would amount to killing the patient to cure a symptom.

Rather than offsetting the direct impact of higher oil prices, the main argument in favor of hiking rates in a stagflationary environment is secondary effects. The idea is that restrictive policy helps ward off the dreaded wage-price inflation spiral, where rising prices on consumer goods lead to higher wages which in turn set off another round of consumer inflation. It was this type of self-reinforcing mechanism that played a key role in U.S. inflation in the 1970s. While this is a nice argument in theory and has a reasonable historical hook, it doesn’t seem to us to have much relevance to modern reality.

In the 1970s, roughly one-third of U.S. workers were unionized, and cost-of-living adjustments were built into collective bargaining agreements. Today, just over 10 percent of workers are unionized, and wages linked to inflation are largely a thing of the past.

More broadly, the rise of independent contractors and gig workers, combined with the threat of AI and the signs of a softening labor market, mean workers have little effective negotiating power to turn higher consumer prices into higher wages. Even during the post-pandemic labor shortages, real median wages in the U.S. only rose $3.00 per week between Q1 2022 and Q1 2024. With labor slack building and the threat of AI overhanging the labor force, is it credible to think that the broad core of workers is going to be negotiating better deals now?

The bottom line, we believe, is that the Fed can make a symbolic move by shifting rates half a point lower or higher, but that should be taken for what it is: a symbol. The real-world economic impact of that type of policy is likely, we believe, to be dwarfed by developments in the Middle East.

The pAIn trade

AI has begun displacing knowledge workers and, in our view, will almost certainly continue to do so. In economic terms, this is a win. The discipline is, after all, about maximizing output given scarce resources. If AI can achieve the same production while using fewer workers, we think an economic boon will have been achieved.

Key to achieving that win is the belief that these freed-up workers will be deployed elsewhere and create new products. On a macro scale, that’s probably true.

But for the newly freed-up worker, the picture is less rosy. Many of them have spent years or even decades acquiring the education and experience that made them “knowledge workers.” Stripped of that value, their employment alternatives will likely be fewer and worse. For those who remain employed in their original profession, we expect AI to leverage their skills and provide an economic benefit. But for the rest, uncertainty reigns.

When we look at the AI-labor dynamic, it seems fairly obvious to us that the forces involved are unlikely to respond to a quarter-point change or two in overnight interbank borrowing rates. It’s not like anyone will keep a human developer on the payroll at six figures instead of a free AI tool just because their payroll costs can be funded at three percent instead of 3.5 percent. Instead, we think fiscal and legislative policy together are the appropriate tools for addressing the economic and social implications of a technology revolution.

Wrong tool at the right time

“Don’t fight the Fed” is an aphorism for a reason. Within its sandbox, the institution is incredibly powerful. But when it comes to issues such as supply shocks and labor markets upended by innovation, we think the limits of the institution’s power become apparent. The Fed will no doubt do what it can, but what it can do may fall far short of what is needed.

Categories

Analysis