The risk of government bonds

The problem with such assumptions is only really a problem when those assumptions are enshrined in laws and regulations. One example I have frequently devoted a column to in recent years was the end of the neutral portfolio. For years, the neutral portfolio (50/50 or 60/40) was a much better investment than just shoving all your assets into the stock market, but because risk profiles are laid down in laws and regulations, at a certain point people only invested in (government) bonds because the same laws and regulations required it, something that simply cannot be changed quickly. Of course, there were possibilities to deviate even then, but in supervisory land, it is risky to stick your head out of the sand.

If laws and regulations go wrong, they can quickly create systemic failures and thus systemic risks. They usually come to light in times of waning liquidity when Warren Buffett goes to see who is wearing swimming trunks. A great - now classic - example was the Basel regulation of banks a decade ago. Because government bonds from the European economic area all had a 0 percent (!) risk weighting - whether they were German or Greek - it was attractive for banks to buy them. That caused a strong convergence of interest rates in the eurozone, but thus also much bigger problems when Greece (and with it many commercial banks) threatened to collapse. Pension funds are also facing much larger than necessary losses (on government bonds) in 2022. The large positions in negative-yielding bonds were there mainly because of long-term liabilities. These long-term liabilities had to be covered with (government) bonds because they were considered risk-free. In fact, in terms of duration, equities were a much better hedge for these long-term risks.

Also in the United States, the percentage of equity that banks must hold for a position in Treasuries was not high. From a credit risk perspective, this is justified, but not from a duration risk perspective. Moreover, overall duration risk has increased sharply in recent years because of the sharp decline in interest rates. When interest rates rose last year, the losses came, losses far greater than were statistically thought possible (again). By the way, the losses in the U.S. banks that have fallen are limited. In that context, fully guaranteeing all deposits is a very heavy remedy. If depositors had been given a haircut of 10 percent that would probably have been enough.

The annoying thing about the government bond asset class is that this is not a small category and measures have already been taken to secure government financing. The fact that the central bank is monetarily financing the government is in itself a reason to warn of long-term inflation, but in the short term, the current turmoil could cause even more problems. Virtually every bank has government bonds on its balance sheet and with them, the paper losses incurred last year. In that regard, it is not a good sign that Credit Suisse was still struggling to keep credit lines open with other banks after a 50 billion support package. Meanwhile, 1 billion of that is UBS's problem.

The robust intervention of regulators also suggests that they see much bigger problems as well. With the fourth bank to collapse, the market is beginning to disbelieve the repeated statement that this too would be a unique problem. That statement now acts rather like a red rag on a bull. What the central bank actually wants to say is that even for central bankers, the problems are much bigger than they expected. Consequently, there is no bank with as many government bonds on its balance sheet as the central bank. So there are huge losses on those. It will cause no dividends to be paid for the time being.

Rather than classifying government bonds as risk-free, one should rather speak of unlimited risk. This is because there is virtually no bank that does not have a (relatively large) position in government bonds. The moment these banks are forced to take less risk, suddenly everyone is on the sell side and liquidity rapidly decreases. And government bonds are found in more places in the economy. For example, they are used as collateral in financial agreements. There are already corporate bonds that give a lower interest rate than comparable government bonds, something that, by the way, was very common 50 years ago. If collateral in the form of government bonds can no longer be trusted then it can cause a lot of contagion through the derivatives market. Indeed, counterparty risk is always there in the financial system. Only hard assets (without counterparty risk) are then still an alternative.