How to Fix the Central Bank
#108, May 2025
Often I draw on my experience writing succinctly in Twitter to inform my newsletter message. This week (and because I should have sent this out yesterday) I am doing things in reverse, so the messages are a bit scattershot tonight. Spanish-readers can see the Twitter version soon, probably on Saturday.
As the BCRD has just shown with another year’s audited accounts, it has a big hole in its pockets, with perhaps a percent of GDP’s worth of money vanishing each year, and a negative net worth that is either already, or soon will be (think about Dr. Evil from Austin Powers) ONE BILLION PESOS (mwah-hah-hah!).
It makes a good joke, but it sits heavily on the shoulders of the Dominican people. And although the central bank has elected policies that make the problem worse, there actually ARE solutions. As you might imagine, none of them are fun—and some are terrible—but they are potential solutions to a problem at an institution that extorts money out of the limited government budget each year, under risk of starting a death spiral for the central bank (where debt is issued to cover interest costs that are ever-higher).
Possible solutions basically fall into three categories: make the taxpayer pay; make the banking system pay; and restructure the BCRD balance sheet. Let’s look at all three.
1. The simplest solution is just to recapitalize the bank with taxpayer money. This was the basis of the 2007 recapitalization law (that was never fully implemented), and is the basis of the IMF’s recommended solution. They had a team look at the problem in 2023, and it came up with the follow set of annual transfers (the red line being the current practice):
The IMF talks about issuing various sorts of government bonds, but basically amounts to doubling current budget transfers, which were RD$46 billion last year (0.6% of GDP). These recommendations are reprised on pp. 85-88 of the most recent IMF report: https://www.imf.org/en/Publications/CR/Issues/2024/09/12/Dominican-Republic-2024-Article-IV-Consultation-Press-Release-and-Staff-Report-554787 In its favor, this approach is the clearest and most straightforward. However, with a Tax/GDP ratio of only 15%, using 1% per year of taxpayer money for the central bank is, um, not the most equitable solution. More important, though, is that the numbers here start from the BCRD’s own accounting, which treats most of the BCRD’s negative equity as an asset, because the government owes it. Still, the chart is useful as a way to show what it would take for the trend to start moving in the right direction.
2. A second solution is to make the banks pay. Inevitably, the bankers will make their customers pay, but, as the problem was begun by the 2003-04 banking crisis, there is a certain justice in beginning there. How would that work?
Up until May 2009, banks were required to hold non-interest bearing deposits at the BCRD equal to 20% of their deposits (see Dominicanomics #77 from last November). Over the next 4-5 years or so, this was brought down to around 12%, though the BCRD has not been transparent with the public about the actual rates. With deposits of around RD$2 trillion subject to reserve requirements moving to a 20% unremunerated required deposit would mean RD$160 billion more in required reserves, which would allow the redemption of a similar amount of central bank paper—and perhaps RD$15 billion per year (0.2% of GDP) in interest savings. Banks would likely try to increase lending margins to recoup this, but passthrough would likely not be complete.
In addition, it is reported that the government keeps substantial deposits at state-owned Banco de Reservas. Requiring those deposits to be held at the central bank instead would allow further runoff of BCRD securities, once again reducing interest costs.
Finally, as the banking crisis at the origin of the central bank’s problems was caused by an ad hoc decision to guarantee all deposits, the banks now all benefit from implicit government deposit guarantees. It would not be unreasonable to impose a tax on the banks to recover some of the past losses and to compensate from the implicit insurance currently being provided.
3. The final solution would require a major shift in monetary and exchange rate policy. The BCRD’s net international reserves at the end of April were over $15 billion, equal to about 12% of GDP. This is an enormous sum. While the IMF argues for even higher reserves, it seldom discusses the cost of holding reserve balances, which in this case is substantial. If the BCRD sold all of its reserves at current exchange rates, it could pay off all of its outstanding debt. Problem solved! In reality, though, the rate would not stay the same, and the reserves do serve a useful function as a backstop for exchange rate management and to give confidence to foreign investors that the central bank has the dollars to take their money out of the country on demand. The narrow interest rate spread between U.S. Treasuries and Dominican bonds—little more than 2% right now—despite the country’s bonds being two grades away from investment status has a lot to do with those $15 billion that is just parked. In 2024, the central bank earned RD$33 billion, or about 4%, on those reserves, while paying, let us say, 10%.
Consider two possibilities. First, a gradual reduction of international reserves of, say, $5-6 billion might save another RD$200 billion per year, another 0.25% of GDP annually. There might be some offset in terms of higher borrowing costs in international markets, but that is a problem for fiscal policy—the budget deficit should probably come down from the current 3% of GDP annually.
The second, and more unorthodox solution, would be to sell off all of the foreign exchange reserves and commit to a freely-floating exchange rate. This is not to say the authorities would have no say, but that they would need to use monetary policy as their main tool of exchange rate policy: tightening policy when the peso seemed weak, and loosening when the exchange rate seemed strong. This would certainly entail increased volatility for a while, but this would likely settle down with experience, and with a transparent and credible central bank that coordinated policy with the Ministry of Finance. On $15 billion of foreign exchange reserves, a 6% annual saving would translate into about RD$50 billion per year in interest savings, equal to0.6% of GDP.
The giant hole in the BCRD balance sheet is not a problem as long as the net worth starts moving in the right direction. The three approaches I have discussed are probably best thought of as a men of options, to be combined in some way to put the BCRD on the right track at last. Current transfers are basically enough to maintain the status quo. Some combination of the one-off balance sheet measures that would allow a drop in the debt would go a long way to restoring confidence in the BCRD, though. There a re difficult discussions needed to put together a convincing package that does not overly rely on taxpayer subsidies, but these decisions need to take place, I would argue, and sooner rather than later.