While there are prescriptive upshots to this piece, this piece is meant to get at the core descriptive points about how monetary policy works. There is plenty of room to debate optimal policy and optimal macroeconomic objectives, but we should first be clear about what the Fed's approach to changing interest rates involves.
The Fed is a powerful institution and monetary policy is a potentially powerful lever. By monetary policy, let’s be more precise: the Fed uses its micro policy instruments to exert strict control over short-term (‘risk-free’) interest rates. The Fed’s ability to change the Federal Funds Rate — the overnight rate at which banks lend reserves to each other — has coincidental and reliable control over virtually all money market interest rates. These changes have a number of potential downstream impacts on the financial system and the economy as a whole. But as you travel downstream, the Fed’s power comes with increasingly less control. And when it comes to affecting how much Americans are willing to pay for their consumption goods and services, the Fed’s influence is very attenuated and far from reliable.
Power is about the average and maximal scope for impact. Control is about precision and predictability. So when Larry Summers matter-of-factly told Jon Stewart “The Fed has control of the overall level of demand”, he is simply not correct. No matter what you think the Fed should be doing with interest rates right now, the Fed’s interest rate changes cannot be described as having predictable and precise effects on demand. Nor is it clear that these effects are exclusive to demand; the Fed’s policies also have substantial influence on the supply side. The Silicon Valley Bank failure should be a wake-up call that ever-shifting monetary transmission mechanisms matter.
There are three linkages that are highly critical to understanding how the Fed affects consumers’ level of demand and inflation as a result:
- The effect of rate hikes on broader financial conditions.
- The direct effect of financial conditions on real economic decisions, especially to spend on labor and capital formation, and to a much lesser extent the direct effects on consumption and prices.
- The effect (of higher labor income growth) on households’ demand for consumption and inflation
The Effect of Rate Hikes on Broader Financial Conditions.
The effect of rate hikes on broader financial conditions is subject to time-varying nonlinearities; in plain English, the effect is unreliable. Just because you know that the Fed is hiking or cutting does not mean you can calculate every asset price in the world. The key assets of relevance to the real economy are the costs of debt and equity capital. Those assets are partially determined by the interest rates the Fed sets, but they are also set by growth expectations as well as fickle risk premiums (the expected return earned in excess of the risk-free rate of comparable duration).
Credit and equity risk premiums are highly relevant in the context of spending activity and recessions, and yet the Fed’s effect on these risk premiums is unclear. We know there is some relationship, but it's far from a scientific law. The speed with which the Fed raises interest rates adds to interest rate volatility, which likely does push up these risk premiums. And when we see signs of uncertainty about the growth trajectory reflected in these risk premiums, we also see how the Fed pivoting to a more accommodative path can calm down these same risk premiums. But the Fed does not have mechanical control here, and thus corporate borrowing costs and equity prices can still substantially vary. The Fed can try to offset these (partially) exogenous dynamics through altering their projected path for the Federal Funds Rate, but the Fed is ultimately not in full control here.
The Direct Effect of Financial Conditions on Real Economic Decisions
When Financial Conditions Motivate Spending
The direct effects of financial conditions on real economic decisions remain substantially under-researched. The most obvious and understood linkage here is between mortgage rates and homebuilding. While the total number of housing permits and starts is determined by a variety of supply and demand factors, interest rate hikes and cuts do seem to accelerate and decelerate homebuilding activity. This makes sense given the unique availability of fixed-rate mortgages in the United States. When rates are lowered, homebuilders are marginally more confident they can build and sell residual inventory; when rates move higher, homebuilders fear that any residual housing inventory they have built will remain unsold or take a steep discount (as occurred in the late 2000s). Of course, builders pulling back will ultimately have some depressing effect on the future supply of housing, which is...awkward.
The story is somewhat similar for residential construction employment, though even here, there are nuances of relevance. Lead times for completing building units have blown out over the past two weeks, in part due to physical input shortages and due to labor procurement challenges. Moreover, builders’ order books may already be full and in backlog (past demand that translates into future supply). In which case residential construction employment might be relatively insulated to the initial fall in building permits and start, or at least see a longer lag.
The nexus between financial conditions and business’ expenditures (on capital and labor) are not as reliable but likely underrated given their potential power. CEOs and CFOs look at their balance sheet and their (estimated) share price when setting their budgets. They are typically looking to hit specific return on equity (ROE) goals. It is therefore not that big of a surprise to see so much attention paid to tech layoffs, where valuations rested on speculative risk premiums and low interest rate assumptions. Likewise, when US energy producers were using the high yield energy market as a major source of capital, US energy fixed investment was highly correlated with the state of the high yield debt market. The willingness to extend credit and utilize credit has underrated supply-side effects, as the latest research from Alexi Savov, Phillip Schnabl, and Itamar Drechsler carefully document.
It should also be noted that while large drawdowns in riskier asset prices do have informative value for business fixed investment and hiring patterns, it’s far from a mechanical effect. Business fixed investment is mostly subject to sticky hurdle rates and not always super sensitive to changes in their effective financing costs. A lot of other companies saw their share prices take a hit in the past 15 months but continued to scale investment and hiring; these firms likely did so because they saw missed opportunities from prior capacity constraints and supply chain fragilities. The rapid recovery in demand may have also led to overfilled order books and historic backlogs; these factors can provide a unique cushion when deciding to scale up investments.
When Financial Conditions (Don’t) Motivate Pricing
The direct effect of financial conditions on pricing decisions seems much more nebulous now. Following the breakdown of the Bretton Woods System in 1971, the US (and global) economy entered a brave new world of floating exchange rates and their volatility. In August 1978, the US dollar had depreciated versus the Japanese Yen over 30% from a year prior. Given that businesses were still adjusting to the new system and its volatility, it was also more likely that such exchange rate depreciation would filter into consumer prices.
But we live in a different world now; the US dollar is much more stable relative to major currencies and has been appreciating, even before inflation began to surge or the Fed decided to raise interest rates. Nor are there strong channels for exchange rate passthrough to consumer prices (especially if you look granularly from exchange rate to import price component to consumer goods price component). The monetary transmission mechanisms change over time.
No Need To Exaggerate, But There Are Even Direct Inflationary Implications From Tighter Financial Conditions
There are also counterintuitive pricing dynamics that emerge from tighter financial conditions. Higher interest rates have shifted the priorities of firms between growth and free cash flow generation. Up until the failure of SVB, there had been a transition in which firms and investors were being rewarded. In the 2010s, low interest rates motivated a ‘growth-at-all-costs’ strategy that made some intuitive sense and likely took on its own momentum. But since then, we have seen a shift to “value” investing and with it, margin sensitivity. Firms are now more eager to sacrifice volume growth for the sake of preserving and growing operating margins, and greater free cashflow in the process. Investors have raised their discount rates and expect more near-term free cash flow generation. Amazon, Uber, and other ‘tech’ companies have consciously pivoted and are willing to raise prices and sacrifice some volume if it means higher and more stable revenue.
We don’t want to exaggerate this counterintuitive effect; down is not suddenly up. But we do want to raise this point to flag just how awkward a policy tool the Fed wields. If the goal really is to reduce inflation, we might want to design more direct policies and policy tools on both the supply- and demand- side.
The (Secondary) Effects on Consumption and Inflation
If you were unemployed but then find a job, you probably would scale up your consumption. If you lose your job, you probably will scale down your consumption. But if you earn a higher wage, will you increase your consumption proportionally? Or will you try to save more of your increase? That probably depends on how much you already have saved up, your age, and a number of other personal factors.
Households fund most of their critical consumption growth (food, energy, transportation, utilities, rent) through their major source of income. For most households, that’s their paycheck(s). For those who need to go into debt to fund that kind of spending, it’s usually credit card debt, which is subject to extremely high interest rates and relatively invariant to Fed policy.
There are some narrow consumption goods where the level of interest rates can shift demand, through credit availability. But these interest rates are hardly representative of all consumption or even reliable on their own terms. Nor do they necessarily lead to demand-sensitive pricing. Very few people toggle their spending based on the level of asset prices (the wealth effect) and based on recent episodes in which the stock market has sold off meaningfully, the effect continues to underwhelm.
Unlike other countries (like Canada and the UK) which primarily have variable rate mortgages, the US primarily has fixed rate mortgages. That closes off a major pathway through which rate hikes could reduce consumer spending without necessarily reducing employment and wages. Most households in the United States do not carry "Fed-sensitive" floating rate debt.
Almost all roads to lower consumer spending must run through labor if the Fed is to be “successful.” But these effects are hardly the kind that allow us to confidently say “if labor income growth is 5%, consumer spending growth will be 5% and inflation will be 3.5%.” Even here, it pays to be humble about the full scope of reconcilable data constellations. Saving rates can change. A wage increase of an already-employed worker has a different effect on consumption than an unemployed worker finding a job. In terms of its effect on the composition of consumption, a lower wage worker seeing a wage increase hits differently from a higher wage worker seeing the same percentage raise.
Ranking The Fed’s Capacity To Control
If we were to do a quick-and-dirty ranking of the Fed’s ability to control the following variables, we would proceed as follows:
- Short-term interest rates
- Longer-term Treasury Yields / risk-free rates
- Mortgage rates and related securities
- Interest rates on other private loans and securities
- Exchange rates
- Residential Fixed Investment
- Risk premiums and security prices sensitive to such risk premiums
- High yield debt prices / interest rates
- Equity prices / cost of capital
- Credit Risk Premiums (Credit Spreads)
- Liquidity Risk Premiums (the expected excess returns associated with relatively illiquid assets over comparable liquid assets)
- Equity Risk Premiums (the expected excess returns associated with equity securities relative to risk-free rates)
- Net Exports & Business Fixed Investment / Capital Expenditures (Nominal, then Real)
- Labor Income / Payroll Expenditures (Nominal, then Real)
- Consumption (Nominal, then Real)
- Consumer Price Inflation
It may seem harsh to rank inflation last, but if you’re ready to think through the Fed transmission mechanisms likely to be of highest relevance, there is probably less room to quibble. The world changes all the time and so too will our ranking; if we saw evidence of exchange rate depreciation and substantial passthrough risk, we would change our view as any right thinking person should. If the US switched to variable rate mortgages like other countries, that would provide a more direct mechanism for reducing consumption through rate hikes.
If you believe—as we do—that inflation is really the last link in the causal chain of monetary transmission and can easily be swung by non-monetary factors, then it might be time to develop more direct tools, whether from the demand- or supply-side.