Why can sometimes stock prices rise when interest rates rise?

Basic macroeconomics theory states that stock prices are inversely correlated with interest rates, i.e., when interest rates rise, borrowing is more costly, and thus companies with huge debt would be hurt. As a result, their expected earnings would be forecast to decline, which would eventually translate to declining stock prices.

However, history suggests than they are not always negatively correlated. Therefore, under what conditions does the opposite hold, i.e., stock prices are positively correlated with interest rates?

-

In the chart below, I'm showing the rolling correlations between stock returns and bond returns. (The relationship would be flipped if you are studying stock returns vs interest rates).

As you can see, for the bulk of the history since 1960s, bond returns and stock returns were indeed positively correlated; i.e., when stocks went up, bonds went up too (and interest rate declined). The past 15 years or so, however, bond returns and stock returns have turned negative (i.e., stocks go up, bonds go down/interest rate goes up).

The historical "norm" was that negative supply shocks tended to cause the equity market to tank, while the coinciding high inflation hurt bonds as well. Nowadays, with inflation expectation very well anchored, bonds have largely become synonymous with recession-hedges – people buy bonds (cause rates to decline) when stocks are performing poorly ("flight-to-quality").

The past few years' experience is also interesting. The Fed buying up a lot of Treasuries (quantitative easing) plus the zero-interest-rate-policy has emboldened risk taking, sending S&P to records.

So to sum up, how stocks & interest rates move relative to each other is very much dependent on the macro backdrop.

Reference: Antii Ilmanen's "Expected Returns" (2011)

-

If you're more of a Derivatives prices thinker, the drift of the Stock Price under the risk neutral measure is the risk free rate. Which means that the diffusion for the stock price $dS_t = r(S_t, t) dt + \sigma(S_t, t) dW_t$ has a larger drift.

-
A building blocks hypothesis can be similarly appealed to. – user2763361 Jun 25 '14 at 7:47
The risk free drift has nothing to do with the real world drift. It just means that forwards are priced with the risk-free rate. – Richard Oct 9 '14 at 6:37

If the companies asset duration is smaller than the liability duration the correlation with interest rates can be higher than a typical firm. More particularly look for surplus in insurance portfolios and pension plans with a positive duration. If interest rates rises the surplus increases strengthening the BS which will have a more higher price impact.

-

Rising interest rates are sometimes a proxy for rising growth. So if stocks benefit from higher growth more than they are hurt by higher interest rates, stocks will rise.

A statement that "stocks fall when interest rates rise" assumes "all other things being equal." In the first paragraph, that was not the case.

-

Technically, anything can happen: stocks can rise in rising and falling rate environments. There is no rule that stocks can only rise when rates fall.

Other times the market may rise if rates rise less than the expectation

-