Why is investment insensitive to interest rates?

Plenty of evidence suggests that investment doesn't react much to interest rates. How could this be?

Note: any and all opinions are entirely my own and do not necessarily reflect those of the Cleveland Fed, the FOMC, or any other person or entity within the Federal Reserve System. I am speaking exclusively for myself in this post (as well as in and all other posts, comments, and related material). No content whatsoever should be seen to reflect the views of the Federal Reserve System.

---

Happy New Year, my much-appreciated Odds & Ends readers!

This post will start off the year by talking about interest rates. I’m sure there will be much more of that to come, since the Fed is set up to have the ability to raise interest rates as soon as in March (and will likely do so three or four times this year). This post will dive into an interesting finding out of the empirical literature: business investment seems to not care too much about the interest rate.

As bizarre as that might seem at first, there has long-standing been suspicion that investment is interest rate insensitive. Since investment spurs employment and output, this is an important thing to understand if we want to understand how the Fed can achieve the maximum employment side of its mandate, and even the inflation side too.

We’ll explore why and how investment can be insensitive to the interest rate, and a little bit about what that means about the monetary policy process. I tried to keep this post from getting too technical, but I tried to link to relevant papers in case you want to look into this further.

Now, let me make assure you that I didn’t just make this up.

No, I didn’t just make this up

The sensitivity of investment to shifts in the interest rate (or lack thereof) has been questioned by economists for quite some time now. Keynes himself had his doubts; of the two factors that he thought determined the willingness to invest, he considered current and expected demand to be much more important than the interest rate itself. Keynes passed this belief down to many of the OG Keynesians, including Lawrence Klein.

The assumption that investment was largely interest rate insensitive was heavily criticized during the 1970s and early 1980s. The Fed led interest rates to follow a rollercoaster pattern of extreme ups and downs at the time, culminating in a by-choice recession in the early 1980s. Maybe investment could be interest rate sensitive—so long as you’re sending interest rates through the roof.

The modern literature on interest rate sensitivity began toward the end of the 1980s. Early estimates generally found low levels of sensitivity (see Chirinko 1993, Bernanke and Gertler 1995, and Chirinko et al. 1999). Early work emphasized how the availability of alternative financing sources could give firms a lot of wiggle room before interest rates began to matter and searched for other factors in determining investment (e.g., Bernanke 1983, Bertola and Caballero 1993).

Macro insensitivity

We’ll begin by taking a look at investment at the macro level. Measuring investment sensitivity is tricky, perhaps in no small part because investment is simply insensitive.

A recent study, Baldi & Lange (2019), finds a strong shift beginning in the mid 1980s in the sensitivity of investment with respect to the interest rate in the United States, going from quite sensitive to hardly sensitive at all. Several other papers also detect a break in the 1980s, even for employment and inflation’s sensitivity to interest rates. This could explain Bernanke’s (other) 1983 finding that interest rates did matter (but not as much as expected returns), especially considering how opposite his later findings were.

They cite three main causes for this shift. First, changes in financial markets and their regulations may have warped monetary policy transmission; second, globalization has brought not just domestic but foreign interest rates into the picture; and third, the shift from rate-sensitive and capital intensive manufacturing to less-sensitive services has shifted the aggregate level of sensitivity. Caggese & Pérez-Orive (2020) and Döttling & Ratnovski (2020) find that the rise of intangible assets—like software or patents, for example—has lowered aggregate levels of investment sensitivity to interest rates since these assets don’t respond to shifts in the interest rate as much as other capital does.

Sharpe and Suarez (2015) use survey data from CFOs to find that only 8% of firms (!) would increase investment in response to a 100 basis point decline (!) in their borrowing costs; only 16% would respond to a 200 basis point decline! Firms are slightly more sensitive to increases in borrowing costs, but still only 31% would change plans in response to a 200 basis point increase. That should strike you as shocking: the Fed moves interest rates around in 25 basis point intervals these days, but yet look at how few firms would respond to a rate hike eight times the size of a normal hike!

Before we move on, one important thing to remember is that just about nobody borrows at the fed funds rate, which is generally the interest rate being considered in empirical research. The rates that firms actually borrow at can swing further than the shift in the fed funds rate in many circumstances, which can make those 200 basis point shifts in borrowing costs more possible (though still unlikely). In theory, the fed funds rate is often expected to move these rates too; in practice, this isn’t at all guaranteed. We’ll consider the implications of this further in the conclusion.

Why so insensitive?

By looking at the micro level, we can understand why investment is insensitive. Sharpe and Suarez’s survey data suggest that this insensitivity can largely be attributed to the ability to finance investment “internally” by using cash savings or preexisting cash flows (consistent with findings in other work) and the fact that firms seem to set strangely high “hurdle” rates of required return. Most simply, firms make investment decisions based more so on current and expected demand than on interest rates.

Having the option to use one’s own cash savings or cash flows to finance new investment rather than external funds can make interest rates irrelevant. If the cost of using external funds is too large, able firms will simply just use their own finances. Firms might opt for smaller investments when borrowing costs rise too much and they have insufficient cash to cover the full investment, but still the capability to self-finance reduces the effect of interest rates.

Fazzari et al. (1988) is a dated but essential look into how financial constraints affect firm investment. Firms are instead more sensitive to their cash flows, which can help explain why current and expected future demand are important factors. Investment declines a lot, to no surprise, when firms exhaust their cash flow capabilities, but this effect is more muted among older and more mature companies. In fact, evidence from the UK suggests that this gave larger and more established firms a leg up in the recovery from the financial crisis, since smaller and/or weaker firms struggled to make the switch to self-financing as easily.

Sharpe and Suarez (and others) also find that firms seem to set “hurdle rates” (the required expected rate of return in order for a firm to be willing to go through with an investment) that are well above rates that would make sense given the actual costs of borrowing. In fact, the average hurdle rate sat at 14% in 2012, only slightly lower than those reported in the 1980s. This is despite the fact that corporate borrowing costs were only around 4% in 2012 compared to roughly 14% in the early 1980s! The large discrepancy here helps to explain why small shifts in interest rates aren’t going to do much to stimulate investment. Compared to the revenue, the cost of investment in the form of interest is quite small at today’s rates.

It’s not all insensitive

One interesting exception has to due with inventories. Firms can be sensitive to interest rate shifts through an “inventory” channel, as discussed in a fascinating paper by Maccini, Moore, and Schaller. Recall that investment includes more than just the obvious examples that often come to mind, like big manufacturing equipment or a new office building. Businesses’ inventories are investment, too. Firms generally want to keep enough in inventory to avoid going out of stock, but they also face an incentive to not hold too much inventory since the opportunity cost of that “excess” investment is the interest rate.

While some more “raw” analyses of inventories show that they’re quite insensitive to interest rates, the authors of this paper find that this is not the case in a well-designed model. They find significant roles for the interest rate as an opportunity cost in determining inventories. Further, they find that firms do actually build up inventories in response to a decline in the real interest rate, so long as that decline is not expected to be temporary. Perhaps this can help explain why firms only began consistently building up inventory in 2011, two years after the Fed cut rates to zero, when it finally convinced markets that it would keep rates low for longer.

Why do interest rates matter, then?

Well, they do for all kinds of other reasons—we’ll get into that another time. But as it pertains to how interest rates can affect firm investment decisions, we shouldn’t really expect a 25 basis point rate hike or cut to affect investment.

Now, that being said, a 25 basis point rate hike or cut (or even the decision to not change rates at all!) can lead to a much larger change in borrowing costs than just 25 basis points. Pretty much nobody borrows at the fed funds rate; instead, bank loans and market rates like corporate yields are typically spreads above the fed funds rate and Treasury yield, respectively. If a rate decision tightens (or loosens) financial conditions too much, it might increase that spread (or reduce it), leading to a total change in borrowing costs that might be much larger than just 25 basis points.

As observers of the 2008 financial crisis should know, bank and market interest rates are under no requirement to decline just because the Fed is cutting interest rates. True, many firms could (and did) switch to internally financing their investment, but consumer spending cuts drag down businesses’ cash flows too, making self-financing more difficult. Further, even if a firm still didn’t care about the interest rate, monetary policy decisions still can determine the willingness and capability for banks to even offer credit.

Monetary policy decisions ease and tighten financial conditions and determine risk premiums faced by borrowers. The interest rate can do this, but think back to my last few posts about how the Fed affects the whole yield curve by setting expectations about the future; credible forward guidance can shift financial conditions this way, too. Further, Fed decisions affect the prices of all kinds of other assets, which can also affect firms willingness and desire to invest.

Therefore, there remains a large role for monetary policy. Instead of expecting monetary policy decisions themselves to affect firm behavior, we should look at how we expect monetary policy decisions to affect financial conditions. But even then, in dire circumstances it might very well be the case that Keynesian-style fiscal policy is needed in order to stimulate higher levels of investment.