Covered Bonds
There's been some discussion on the Mule Stable about covered bonds and who may or may not be disadvantaged by them, so I thought I'd post a (very) simplified example of how they would work compared to the alternatives of standard financing or securitisation.
The first (sometimes confusing) thing to remember is that loans are a banks assets and deposits are their liabilities, as are their other sources of funds such as bonds issued by the banks. Imagine a bank has $100 million of assets and, to make life easy, we'll assume they're all home loans. Bank capital requirements mean that the bank has a minimum capital requirement of 8% of "risk-weighted assets". I'll ignore the technicalities of risk-weighting and assume that the home loans are 100% risk-weighted (in fact they'd be less that this, but other loans would have higher weights), which means that the bank must have at least $8 million of capital. Again, in the interests of simplicity, I'll assume that this is all in the form of equity. The remaining $92 million I'll assume to be made up of $52 million in deposits and $40 million raised by issuing bonds which have been bought by superannuation funds (i.e. pension funds).
So, here's the bank's balance sheet:
Assets
Home Loans $100m
Liabilities
Deposits $52m
Wholesale borrowing (bonds) $40m
Equity
Shares $8m
In a moment, I'm going to imagine that the property market collapses and the portfolio of home loans loses 50% of its value. But before that, $5m of the bonds are due to be repaid and the bank has to decide how to refinance them. The options being considered are
- Issue $5m more bonds
- Issue $5m in covered bonds, pledging $5m of the home loans as collateral
- Securitising $5m of the home loans.
In the first case, the balance sheet ends up the same as before. In the second case, the balance sheet looks the same, but one should note that $5m of the assets are effectively tied up for the exclusive benefit of the investors in the covered bonds. In the third case, $5m of assets and liabilities are taken of the bank's balance sheet. This means that the capital requirements for the bank have also dropped marginally as they now only have $95m in assets so only need capital of $7.6m. Since this is a bank that likes to be efficient, let's assume they take advantage of this by buying back $0.4 million of the shares, funded by an additional wholesale borrowing of $0.4m. So, in that case the balance sheet now looks like this:
Assets
Home Loans $95m
Liabilities
Deposits $52m
Wholesale borrowing (bonds) $35.4m
Equity
Shares $7.6m
Now disaster strikes and the loan porfolio halves in value across the board. What happens in each case? Keep in mind that depositors rank first in their claims on the bank, so they are the last to lose money, except where assets have been explicitly pledged as in the case of covered bonds.
Case 1. The loan portfolio falls to $50m. Shareholders lose everything, wholesale borrowers lose everything and the depositors lose $2m, a 4% loss.
Case 2. Again, the loan portfolio falls to $50 and shareholders lose everything. This time, however, the covered bond investors have a claim on the loans securing their bonds, which are now worth $2.5m and so they lose 50% of their investment, while the other bond holders still lose everything. Since the depositors don't have the benefit of the $2.5m that was pledged to the covered bond investors, they now lose $4.5m or 9%.
Case 3. The outcome is the same as case 2.
So, in this scenario, the outcome is the same whether the bank issued covered bonds or securitised. Depositors would be better off if the bank were to simply borrow from the wholesale markets for their funding. Both covered bond funding and securitisation leave them worse off.
Now imagine a slightly more modest collapse in the portfolio value. This time, they lose 20% of their value. Running through the scenarios again we have
Note: calculations here have been corrected (thanks @Baz_man!)
Case 1. The loan portfolio falls to $80m. Shareholders lose everything. wholesale borrowers lose $12m (=$20m - $8m) or 30% of the $40m and the depositors lose nothing.
Case 2. Again, the loan portfolio falls to $80 and shareholders lose everything. This time, however, the covered bond investors have a claim on the loans securing their bonds, which are now worth $4m and they join the remaining senior creditors to get some of the remaining $1m back. After paying out $52m to depositors (who lose nothing) and $4m to the covered bond holders, there is $24m in assets to apportion to claimants totalling $36m (the $35m senior bondholders and the $1m claim by the covered bond investors). This is a recovery rate of 66.7%, or a loss of 33.3%.
Case 3. This time the outcome is a little different to case 2 as the equity is smaller. Shareholders still lose everything and the covered bond holders still lose 20%, but now the other bond holders lose $11.4m (=$20m - $7.6m - $1m) which is 32.6% and the depositors lose nothing.
So the outcome this time is a marginally worse for other bondholders when the bank goes down the route of covered bonds rather than securitisation.
If there's so little in it, this raises the question why has the regulator (APRA) been happy in the past to allow securitisation but not covered bonds? I believe it is a matter of transparency. In the case of securitisation, if you look at the bank's balance sheet, it's clear that there are only assets worth $95m. When the bank uses covered bond, it appears as though depositors have the benefit of a full $100m, when in fact $5m is pledged to the covered bond investors. While a bank would have to make a note to their accounts to this effect, it may be harder to people to understand.
