THE most dangerous words in finance are: “This time is different.” But sometimes markets can genuinely change. After the 2007-08 financial crisis, markets are less efficient and liquid than before.

The evidence can be found in the currency markets, as a paper* in the latest quarterly bulletin from the Bank for International Settlements (BIS) explains. In foreign-exchange markets it is possible to buy currency at today’s rate (the spot market) or at some future point (the forward market). Any student of the currency markets will quickly come across the idea of “covered-interest parity”. This states that the gap between the spot price and the forward price will equal the interest-rate differential between the two countries.

Imagine that American 12-month interest rates are 10% and Japanese rates are 5%. Japanese investors will be tempted to buy dollars, earn interest on them for a year and then cover the exchange-rate risk through a forward deal. So lots of people will be selling dollars in the forward market. They will keep doing so until the dollar is 5% cheaper there than in the spot market, and there is no profit in the trade.

In the foreign-exchange market, which is highly liquid, the possibility of profitable arbitrage should be rare—the equivalent of $100 notes lying on the pavement. But the covered-interest parity rule has been consistently breached in some corners of finance since 2008. In the immediate aftermath of the collapse of Lehman Brothers, the anomaly could be put down to a temporary freezing of markets. Yet the world is not in crisis mode today.

The BIS argues that two factors explain the phenomenon. First, many participants in the foreign-exchange markets are seeking to hedge their exposures, almost regardless of the costs.

Take a Dutch pension fund which decides to invest in Treasury bonds because it trusts the American government’s creditworthiness. The pension fund’s liabilities—payments to Dutch retirees—are in euros and it does not want to take the currency risk of owning dollars. So it will borrow dollars (in order to buy the bonds) and exchange them for euros in the swap market, the equivalent of doing a forward currency deal.

Another group of inveterate hedgers are international banks which, by the nature of their business, will have both assets and liabilities in a wide range of currencies. When those assets and liabilities are not matched, they will want to eliminate the foreign-exchange risk.

If hedging demand was evenly balanced between currencies, this would not be a great problem.

But it seems there is more demand to hedge American-dollar risks or exposures, relative to the yen and the euro, than the other way round. (The reverse is true for Australian dollars, as the chart shows.)

The effect is to drive up the cost of dollar borrowing in the foreign-exchange swap market, to a point where it is out of line with the cost of borrowing dollars in the money markets. Or to express the problem in a different way, the forward currency rate gets out of line with the interest-rate differential between the two currencies (as conventionally measured in money markets).

At this point, if theory held, the arbitrageurs should swoop in and eliminate the discrepancy.

Either the banks could do this themselves (via their trading desks) or they could lend money to hedge funds that hoped to profit from the anomaly. But in the post-2008 world, banks are constrained in the way they can use their balance-sheets. Regulators have insisted that banks hold more capital to reflect the risks involved in arbitrage activities.

The financial sector will not collapse because covered-interest parity no longer applies. But it is a sign of the times: similar oddities have emerged in the interest-rate swap market. For the efficient-market hypothesis to hold true, markets must be liquid enough for arbitrageurs to bring prices back to normal when anomalies occur. But banks are unable to provide the same levels of liquidity as they did in the past. In a sense, that is a good thing. Banks were not charging enough for the use of their balance-sheets before 2008 and many got into trouble as a result.

But it is also a bad omen for when the next crisis hits. Markets may freeze even more quickly than before and asset prices may get even more out of whack than they did in 2008. As long as central banks are still pumping liquidity into the markets, it is tempting not to worry. But they won’t always be so generous.

* “Covered interest parity lost: understanding the cross-currency basis”, BIS Quarterly Review, September 2016