After getting off to a rocky start last month, China’s local government debt swap program is officially underway. As a refresher, here is the situation, in a nutshell:
The idea is to swap existing high-interest loans — which are a consequence of localities skirting debt issuance limits by tapping shadow banking conduits for cash — for standard muni bonds which will carry yields that are more inline with the supposed credit-worthiness of the issuer. This sounded great on paper, but when the provincial early adopters tested the waters they discovered that bank demand for the new bonds was tepid, leaving the PBoC with two options: 1) buy the bonds outright, 2) create demand by allowing banks who purchases the bonds to pledge them for long-term cash loans. Option number one would simply constitute Chinese QE, while option number two is akin to ECB LTROs and in either case, it gives the PBoC an excuse to implement a large-scale easing program and in the case of the latter option, the hope is that banks will use the cash to lend to the broader economy thus kickstarting growth.
China chose the latter option (for now). On Monday, Jiangsu Province sold 3-year bonds at 2.94%, 5-yr bonds at 3.12%, 7-year bonds at 3.41%, and 10-year bonds at 3.41%. Broadly speaking, borrowing costs were lower than expected, a relief for Jiangsu which pulled an offering in April after bank demand proved tepid. That event effectually forced the PBoC’s hand when it came to allowing purchasing banks to pledge the new issues as collateral for cash loans. In other words, we can thank the failed Jiangsu offering for Chinese LTROs.
Here’s FT with more:
Jiangsu is one of China’s richest and best-managed provinces, but its initial plan in late April to sell Rmb64.8bn of bonds to pay off existing debt failed because state-owned banks balked at an interest rate that was considered too low for the risk involved in lending to the province.
After initially saying the Rmb1tn bond programme would be driven by the market, Beijing changed its mind and issued an administrative order to banks to buy the bonds.
The central government also capped the interest rate that could be offered at no more than 30 per cent above Treasury yields, allowed banks to use the new bonds as collateral from the central bank and lowered benchmark interest rates to make the bonds more attractive.
For reference, here’s a breakdown of debt by region (as a percentage of GDP) and also of regional revenue growth:
While it now appears that the PBoC's support (and heavy hand) will be enough to ensure that the debt swap program will be generally successful at least in the narrow sense of saving local governments billions in interest expense, two things are as yet unclear: 1) what impact will participation in this multi-trillion yuan experiment have on banks?, and, more importantly, 2) will a new directive that encourages local governments to continue to tap LGFV even as the bond swap program is barely off the ground serve to undercut the whole endeavour by encourage localities to accumulate still more high interest debt?
Here's Citi on the first question:
We believe local government bond issuance, despite the low yields, is overall positive for banks. We see the following implications:
Negative interest income impact due to the lower yields on LG bonds vs. LGFV loans. We reckon this yield differential could be 200-250bps lower in this case (assuming LGFV loans at benchmark lending rate). On the present announced Rmb1trn debt swap target, we estimate the interest income loss will be about 1% of industry-wide earnings.
Positive for credit risk because provincial government is surely a better credit risk than an LGFV.
Positive for capital because the risk-weight for provincial government bonds is 20%, much lower than the 100% risk-weight for a corporate loan (under the standardized approach).
Lowers LDR and releases lending capacity. To the extent that there is loan demand, banks can use this extra loan capacity to make up for the loss in interest income.
Consider the bolded passage there and then consider this from SocGen:
If we are right about PBoC’s intention of helping local government debt restructuring, the total size of this programme may match the total size of local government’s debt stock at the moment. Considering that issuance for the fiscal spending in the coming years may also need some help on attracting demand, we would not be surprised by an eventual size of CNY20tn.
The question then appears to be this: if the program is expanded dramatically over the course of the next several years, what will the cumulative impact be on banks' bottom line if just the initial CNY1 trillion pilot program is going to amount to 1% of industry-wide earnings?
As to the relaxation of the ban on LGFV financing, we've suggested that this could ultimately open the door for the limitless expansion of credit in China (and maybe that's the goal) because should the PBoC decide that new LGFV loans are also eligible for the debt swap program, it isn't clear what keeps this from turning into a debt creation machine, whereby local goverments obtain financing wherever they can get it, swap the loans for muni bonds, sell the muni bonds to banks who then pledge them to the PBoC for cash that's then used to extend still more credit.
Irrespective of whether this is exactly how the situation plays out, one thing seems clear: with the relaxation of the LGFV rule, China is sending a clear message that the immediate concern is simply to roll-over local governments' existing debt while allowing them to add still more leverage. In other words, "delay-and-pray." As it turns out, Fitch agrees.
Chinese government directives last week concerning local government debt signal a potentially significant policy shift to prioritise growth over managing the country's debt problem…
Uncertainty over the scale and strategy to resolve high local government debt remains a key issue for China's sovereign credit profile, and the latest directives could reflect a continuation of an "extend and pretend" approach to the issue…
A joint directive from the Chinese finance ministry, central bank and financial regulator on 15 May, instructed the banks to continue extending loans to local government financing vehicles (LGFV)s for existing projects that had commenced prior to end-2014, and to renegotiate debt where necessary to ensure project completion. This is an explicit form of regulatory forbearance, and serves to delay plans to wind down the role of LGFVs. More broadly, it also suggests that propping up growth in the short term has temporarily taken priority over efforts to resolve solvency problems at the local government level.
We'll close with the following passage from Citi which suggests that in reality, the new local government bonds might as well be treasury bonds both in terms of yield and in terms who will ultimately be responsible if (perhaps more appropriately "when" given what we've said above) local governments can no longer kick the can and wind up unable to pay.
The new setting is in line with our view of burden sharing on local debts: local governments have promised no default, banks will receive lower yield, and the central bank has committed to provide cheap funding. But if there is no efficiency gain in coming years, some local governments may become insolvent, and then all burdens would be channeled up to the central government. Local bonds are thus not much differentiated from treasury bonds.