Indian corporate bonds need a buyer and not a lender

The Reserve Bank of India (RBI) announced yesterday that it is setting up a Rs 500 billion Special Liquidity Facility for Mutual Funds (SLF-MF). This is basically a refinance window for banks that lend to mutual funds to help them handle redemption pressures in an environment where corporate bonds are both stressed and illiquid.

I believe that SLF-MF does not solve the real problem at all, because mutual funds by their very design need to liquidate assets to meet redemption. Unlike banks and hedge funds, mutual funds are not designed to use leverage: the Securities and Exchange Board of India (SEBI) Mutual Fund Regulations state:

The mutual fund shall not borrow except to meet temporary liquidity needs of the mutual funds for the purpose of repurchase, redemption of units or payment of interest or dividend to the unitholders:

Provided that the mutual fund shall not borrow more than 20 per cent of the net asset of the scheme and the duration of such a borrowing shall not exceed a period of six months.

Some mutual funds do talk and act as if the 20% limit allows them to borrow to juice up their returns or to speculate on prices rising in future. But the regulations are clear that this is not the intention, and any mutual fund that borrows for such speculative purposes is actually running a hedge fund in disguise. The proper use of borrowing is to deal with operational timing mismatches where a fund is not able to sell assets and realize the proceeds in time to meet the redemption needs.

Back in 2008, when the RBI launched a facility similar to SLF-MF during the Global Financial Crisis, I explained why mutual funds cannot borrow their way out of redemption trouble:

A mutual fund is very different from a bank. When a bank borrows to repay depositors, there is a capital cushion that can take losses on the assets side. When this capital is gone, the bank also needs to be recapitalized and cannot solve its problems by borrowing from the central bank. A mutual fund does not have any capital separate from the unit holders. This means that the only prudent way for a mutual fund to repay unit holders is by selling assets. If it borrows, then it is exposing remaining unit holders to leveraged losses.

The problem that the mutual fund industry faces today is in many ways worse than in 2008. Until Covid-19, open end debt mutual funds could offer redemption on demand to their investors because there was a liquid market for the bonds that these funds held in their portfolio. Now, the assets have become illiquid and hard to value while the investors are still able to demand instant liquidity. This mismatch can be solved only by some combination of two things: (a) the liquidity of the assets could be improved by a market maker of last resort, or (b) redemption could be restricted. The SLF-MF does neither of these, and merely postpones the problem till the maturity date of the borrowing.

Any financial crisis is ultimately resolved by allocating and absorbing losses. Everything else is a stopgap arrangement that merely postpones the day of reckoning. In times of crisis, there is sometimes merit in such postponement because it buys time for a more orderly loss allocation. But, whenever we kick the can down the road, we must ensure that by doing so we do not make matters worse in terms of making the losses bigger or making the ultimate loss allocation less fair or more difficult. The SLF-MF does not pass this test because it rewards those who redeem and penalizes those that remain in the fund (thereby incentivizing a run on all debt mutual funds):

• Those who choose not to redeem will find their mutual funds morph into leveraged hedge funds. (Some mutual funds are already seeking a relaxation of the 20% borrowing limit to 30%).
• Freed from the mark to market discipline of selling assets at market prices, mutual fund net asset values (NAVs) could become increasingly untethered from reality. In all likelihood, those who redeem will exit at an inflated valuation, thereby inflicting losses on those that remain. Sophisticated corporate houses and other smart investors who are redeeming today will get a good deal and the unsophisticated retail investors still holding on to their units will be left with all the rotten overvalued assets.

What then are the solutions to the problems of the mutual fund industry today? I will first outline three solutions that I do not recommend:

• Halt all redemption and turn the funds into listed closed end funds that investors can sell on the exchange at whatever price they fetch in a free market. This might work when the problem is confined to a handful of funds, but the problem in Indian debt mutual funds is too big for this solution. Debt mutual funds are now systemically important as the RBI implicitly recognized when it invoked its financial stability mandate to justify the SLF-MF. So this solution is not recommended.

• Provide a sovereign guarantee to all funds. The US did this during the Global Financial Crisis for Money Market Mutual Funds, but this is not a viable solution for funds with high credit risk that are at the epicentre of the crisis in India.

• The sovereign (or central bank) steps in as the buyer of last resort for corporate bonds instead of as a lender of last resort. This is a well established model that has worked very well around the world:

  • Korea implemented this solution in 1999 after the bankruptcy of several large conglomerates (Chaebol) during the Asian Crisis.

  • The European Central Bank (ECB) and the Bank of Japan (BOJ) have been buying corporate bonds for many years now as part of their quantitative easing and they have expanded such buying significantly after Covid-19.

  • The US has created the Primary and Secondary Market Corporate Credit Facility to buy corporate bonds. The US government provided $75 billion of equity for a Special Purpose Vehicle (SPV) which will be able to borrow from the US Fed to buy corporate bonds of up to $750 billion.

  However, in the current situation, this model creates too much of moral hazard. This is because the corporate bond markets were already under stress for business cycle and other reasons even before Covid-19. Mutual funds and their investors who took on credit risk to earn a higher return in good times should not get a free pass when things had soured even without Covid-19. That makes me hesitant to recommend this solution.

My preferred solution is for the mutual fund industry itself to create the buyer of last resort with only limited support from the sovereign. It is guided by the principle that whenever we bail out the financial sector we do so not to help the financiers but to protect the real economy which depends on a well functioning financial sector. In these troubled times, the real economy cannot afford the loss of any source of funding. This is the same principle that guided my proposal earlier this month for a preemptive recapitalization of banks and non bank finance companies.

My proposal is similar to the US Fed’s Secondary Market Corporate Credit Facility mentioned above with one critical difference. Instead of the equity for the SPV coming from the government, it should come from the mutual funds themselves. When investors redeem from a mutual fund, and the fund is not able to sell bonds in the market, it can sell the bonds to the SPV at a fair value as determined by the SPV. The mutual fund will be required to contribute a percentage of the fair value as equity stake in the SPV and will receive only the balance in cash. If we follow the US and require 10% equity for the SPV, then a mutual fund selling bonds with a fair value of 100,000 to the SPV will have to contribute 10,000 towards the equity of the SPV. The SPV will use the equity of 10,000 to support 90,000 of borrowing from the RBI which allows it to pay 90,000 as the cash component of the purchase price to the mutual fund. The equity contribution of the mutual fund to the SPV has to come from the investors of the mutual fund. So an investor redeeming 100,000 from the fund would get 90,000 in cash and get the remaining 10,000 in the form of units representing the equity stake in the SPV.

This means that a large part of the credit risk of the bond remains with the redeeming investors as a whole. If the SPV ultimately realizes only 96% of the fair value of all the bonds that it bought, then its equity will come down to 6% from the original 10%. The redeeming unit holder will have got (a) 90,000 in cash and (b) shares in the SPV worth 10,000 originally, but worth only 6,000 when the SPV is wound down. The redeeming investor ends up with 96% of the original fair value of the bonds which matches the 96% ultimate realized value of the bonds. On the other hand, if the SPV realizes 103% of the fair value, then the original equity rises to 13% and the redeeming investor recovers 103,000 (90,000 in cash plus shares in the SPV worth 13,000).

Let me discuss some possible objections to the proposal:

• There would still be a large sovereign backstop in this arrangement. The central bank would lose money on its loans if the bond loses more than 10% of the (fair value) purchase price. Such a large loss on a diversified portfolio of bonds can arise only when there is a systemic economic collapse, and it is the job of the central bank and the sovereign to prevent such an economic collapse. In any case, nobody other than the sovereign can absorb this tail risk on a large scale.

• The RBI will have to lend directly to the SPV instead of indirectly via the banks (because the banks are in no position to absorb the tail risk of an economic collapse). There is no legal bar on the RBI doing such lending. Section 18(3)(c) of the RBI Act allows the RBI to lend to any person when, in the opinion of the RBI, a special occasion has arisen making it necessary or expedient that such action should be taken for the purpose of regulating credit in the interests of Indian trade, commerce, industry and agriculture. If the current situation does not warrant the use of Section 18, then that section might as well be scrapped.

  Central banks around the world have realized that as the financial sector evolves from a bank dominated system to a market dominated system, the central bank’s financial stability mandate broadens substantially. Supporting systemically important financial markets becomes as important if not more important than supporting systemically important financial institutions. As a consequence of this, the central bank has to be as much of a market maker of last resort as a lender of last resort.

• Both the SPV and the RBI will need expertise to value the bonds and analyse their risks. In the US, the Fed has hired Blackrock to provide these services and has also designed mechanisms to deal with potential conflicts of interest. Similar arrangements should be possible in India as well.

Finally the proposal reflects a realistic evaluation of the current situation:

• The mutual funds under stress today took a conscious decision to take higher credit risk to earn higher returns. In good times, investors in these funds did earn these higher returns commensurate with the risks. There is no justification for any bailout of these investors who took a wrong investment call.

• Setting aside concerns about possible mis-selling, these mutual funds were legitimate products serving legitimate investment needs. The stresses in these funds today are due to the adverse business conditions today (some of which predate Covid-19). Similar stresses are present in the banking system as well, but they are less visible there as banks do not follow mark to market accounting in their banking book.

• India needs a vibrant corporate bond market, and a vibrant mutual fund industry. From a financial stability perspective, it is important to support these systemically.

• The economic impact of Covid-19 is unknown at this point. Only the sovereign balance sheet can absorb the tail risks of such an uncertain magnitude. Fortunately, by absorbing this risk, the sovereign makes catastrophic outcomes less likely, thereby mitigating the risk that it absorbed.

Posted at 7:09 pm IST on Tue, 28 Apr 2020         permanent link

Comments