Arbitraging upward sloping yield curve

Question

I read from various sources that yield curve is normally upward sloping. If that's the case, if we borrow short term and lend long term, won't we always make money on average?

Let's say 1-year interest rate is 1% and 2-year interest rate is 2%. I could borrow for one year and lend for 2 years. After one year I refinance, and as long as the interest rate is below 3%, I will make money. So the only way I could lose money is if one-year interest rate rises above 3%. However interest rates are as likely to go down as it is to go up, and the chance that it rises above 3% is even slimmer.

So here's my argument. If the shape of the yield curve is consistently upward sloping, the short term interest rate has to go up by a significant amount in order for the "short borrowing long lending" strategy to lose money. However interest rates cannot go up forever. On average, it goes down as often as it goes up. Therefore the "short borrowing long lending" strategy will most likely profitable.

I know there are other risks involved, such as credit risk or inflation risk. But are these enough to explain it?

Answer 1

This is what banks have been doing for hundreds of years. They borrow short term (mainly through deposits and interbank lending) and lend long term (e.g. mortgages).

I would not call it arbitrage, as it is not riskless profit.

Apart from credit risk and interest rate risk, there is also liquidity risk. In these type of strategies, the investor has to renew (roll over) the short term debt. What happens if the short term lending markets dry up and the short term debt cannot be rolled over, as indeed happened in 2007-2008?

Here is an idea: What about buying 30 year treasury bonds and financing this trade through the repo market using the treasury bonds as collateral. This eliminates credit risk, right? Moreover, due to very small haircuts on the repo transaction, we could have a higly leveraged position, using a small amount of own capital. Could we then call this strategy an arbitrage strategy?

The answer is no. Liquidity risk is exacerbated when high degrees of leverage are employed. A famous case study showing this point is the Orange County bankruptcy.