This publication is a product of Carnegie China. For more work by Carnegie China, click here.
The People’s Bank of China (PBoC) is determined to rein in what it fears might be a government bond bubble because of its possible disruptive impact on Chinese banks. Yields on long-dated sovereign bonds have been falling since December of last year when, after a few fairly stable months during which China’s ten-year sovereign bond traded between 2.60 and 2.70 percent, bond prices rose sharply, causing yields to drop 10 basis points during the month.
The surge in bond prices continued over 2024. By the end of the first quarter, yields were down a further 15 basis points to 2.30 percent. They dropped another 8 basis points in April before aggressive action by the PBoC and banking regulators drove yields back up to the 2.30 percent level, where (with a lot of regulatory support) they stabilized through mid-June.
But more recently they have once again been on a roller coaster, and in the past week regulators have been especially aggressive. Yields dropped to as low as 2.12 percent in early August before a week of frenzied activity by the regulators drove them back up to 2.22 percent by the beginning of this week, with nearly half of that occurring on Monday. By Wednesday, however, markets had managed once again to reverse a substantial part of the week’s intervention.
This has become a familiar story. For much of 2024, markets have driven up bond prices as banks and investors aggressively bought up long-dated government bonds, after which the regulators have swooped in with a combination of measures to drive bond prices back down. These measures have been fairly heterodox. They include warnings to the smaller banks, outright prohibitions and cancellations of bond purchases, bank investigations, regulatory fines, and what some are calling quantitative contraction—with the PBoC borrowing long-dated government bonds from banks and selling them into the market to drive down prices.
Last week, the measures intensified after bond yields plummeted another 10 basis points in late July. Regulators announced an investigation into four rural commercial banks for allegedly “manipulating sovereign bond prices in the secondary market”. This was widely seen as a warning to smaller banks who, after the large state banks sold long-dated government bonds into the market—encouraged, no doubt, by strong regulatory pressure—took advantage of these sales to buy up the bonds cheaply. This week, at least four brokers took steps to reduce bond trading, helping to drive down trading volume by nearly 50 percent from last week. They, too, were almost certainly pressured to do so by the regulators.
It seems that the regulators are aggressively intervening in the bond market mainly because they are worried that a bond bubble might lead to financial instability, much as rising interest rates in 2022 and 2023—along with the sharp decline in US government bond prices—led to the collapse of Silicon Valley Bank. Here is the Wall Street Journal on the topic earlier this week:
China’s central bank is set to conduct stress tests on financial institutions’ exposure to bond holdings, marking the latest efforts by authorities to rein in a monthslong rally and prevent fallout risks.
The People’s Bank of China said in its quarterly policy report that the planned checks are aimed at preventing risks stemming from potential rate fluctuations in the future that could dent bond prices and cause financial losses for investors.
Paradoxically, at the same time that it has tried to force up yields on long-dated government bonds, the PBoC has also tried to reduce interest rates as part of efforts (not terribly successful, to be sure) to stimulate growth in the economy. It has allowed repo rates, for example, to fall to some of the lowest levels all year. As a recent Bloomberg article explains:
The saga underscores the dilemma Beijing is in: while it has to support the moribund economy by keeping funding costs low, it has to make sure money isn’t so cheap that a bond bubble is formed that jeopardizes financial stability. That’s why the PBoC seems to be sending mixed signals, slashing a string of interest rates in late July right after implying it may sell its debt holdings to cool the rally.
But while the risk of bond price fluctuations is real, the problem is that the huge popularity of long-dated government bonds is caused by continued weakness in the real estate market and the widespread perception that the economy is still slowing. These in turn lead to the reluctance of Chinese businesses to borrow to expand their production and of Chinese households to borrow for mortgages. In July, Chinese banks loans actually contracted (by RMB 77 billion) as Chinese households and businesses repaid more debt than they took out, the first time this has happened since 2005.
This pessimism is also reflected in the stock market, another market that this year has seen heavy intervention by regulators, this time to get prices to rise. But while the regulators had succeeded for a while (the CSI rose from 3179 on February 2 to 3690 on May 20), they were unable to maintain the market, and it has dropped sharply since then (to 3326 by August 12). At the same time, as Bloomberg notes in another article, “share transactions in China shrank to their lowest level in over four years, as a local bond rally hit fever pitch in a weakening economy.”
Banks and other fixed-income investors are not buying long-date government bonds out of perversity, in other words, but because the economy is struggling and better alternatives don’t exist. This raises an obvious question: Will the regulators succeed in preventing banks from piling even more into long-dated government bonds and driving yields down further? Another way of asking this question might be to consider how, if they are indeed prevented from buying the long-dated bonds, Chinese banks can manage their deposit inflows and the weak real demand for their funds.
I think we should at the very least be skeptical about the ability of regulators to push down bond prices and so raise bond yields. Given that the demand is real, to the extent that the regulators are successful in repressing purchases and keeping yields up, we should consider the secondary implications on the rest of the economy. As I see it, there are at least five different ways the economy can react to regulatory attempts to force up yields in long-dated government bonds.
- The best way, of course, is a revival in the economy that creates new loan demand from the real side of the economy. In that case, banks would have profitable, low-risk places in which to invest, and the pressure to park liquidity in government bonds would abate. This is probably not a very likely outcome, at least in the near term, unless Beijing turns to aggressive, demand-side fiscal stimulus.
- The PBoC can continue more aggressively to borrow long-dated government bonds from the large banks and short them to satisfy demand (that is, more quantitative contraction). The problem with this solution is that it is contractionary, and monetary policy is supposed to be expansionary to accommodate the needs of the economy. If the PBoC takes balancing steps, it will raise long-dated yields while steepening the yield curve. In that case not only would lower short-term rates prove potentially self-defeating (by making the carry trade more profitable), but they would crimp the longer-term rise in the household share of GDP by reducing the return to household savings deposits.
- The PBoC can move in the opposite direction by raising short-term interest rates so as to make long-dated government bonds less attractive. This, however, would fly in the face of easing monetary policy so as to make it more accommodative in a slowing economy. What is more, to the extent that it causes growth expectations to slow even further, it would put yet more pressure on banks to scoop up long-dated government bonds.
- The regulators can simply prohibit banks from buying long-dated government bonds. This wouldn’t, however, resolve the more fundamental problem, which is that banks have limited places in which to invest burgeoning liquidity. In that case deflating a potential bubble in the long-dated government bond market might just create another a bubble in some other, riskier asset.
- Banks can try to reduce their liquidity pressure by lowering deposit rates and attracting fewer customer deposits. Besides putting downward pressure on household income, however, a reduction in household deposits will just transfer the liquidity problem from banks to households. Some of this may be transferred into more consumption, but the more likely scenario is that households will look for other ways to deploy their savings. But as households search for places in which to invest their liquidity, they could easily set off further asset price bubbles in ways that would be more difficult for the regulators to control.
There are probably other ways to balance attempts by the regulators to prevent a surge in bank purchases of long-dated government bonds, but as the examples above suggest, except in the case of an unexpected revival in economic growth, they all run into the same inconsistencies. The regulators, in other words, are trying to resolve one of the symptoms of a deeper underlying problem without in any way addressing the underlying problem itself: that a slowing economy is causing households to cut back on spending while at the same time limiting what they can do with the resulting rise in savings.
Unfortunately, in a slowing economy with rising savings, banks have little choice but to invest in low-risk, higher yielding assets. One way or another these savings have to show up somewhere, and the most likely (and safest) place is in long-dated government bonds.
As long as that is the case, real yields will decline, whether or not the regulators approve. And while normally a decline in yields might boost the economy, under current conditions they are unlikely to have a much less positive effect. What is probably safest for the economy is for investors put their liquidity in long-dated government bonds. The alternatives may indeed slow down the decline in government bond yields, but at the cost of a riskier outcome.
