The Government's investments in the residential mortgage backed securities market, which Joshua Gans and I first advocated in March 2008, have proven to be very successful from both a financial and policy perspective.
Based on a media release issued by the Australian Office of Financial Management (AOFM) yesterday, the Commonwealth has invested $8.7 billion in 27 transactions to date. (In total, $16 billion has been committed to the venture by Wayne Swan thus far.) Most encouragingly, 'spreads', which is the market vernacular for the cost of funding that lenders pay to investors, look to have contracted further, with the Commonwealth now willing to purchase AAA-rated home loans at just 1.1 to 1.3 per cent over the benchmark rate (this is known as the 'bank bill swap rate' or BBSW). Private sector demand for these securities is very healthy, and the Commonwealth has been able to scale back its investments to facilitate more private market participation.
When Joshua and I first published our work we estimated that the 'break-even' spread for most securitisers was about 1 per cent over the benchmark rate, which was subsequently confirmed by the RBA. Yet with the expansion in mortgage margins effected by the major banks, combined with a significant reduction in the cost of selling home loans (via big cuts in the commissions paid to brokers), many smaller lenders claim that they can make money at spreads of around 1.3 per cent over the benchmark rate. That is, broadly in line with where the market and the AOFM are independently offering funding today.
How much money are taxpayers making from this policy? We can look at this two ways: intuitively and then by referencing the AOFM's actual reported performance (I only did the latter after I wrote the first iteration of this post!).
Since the cost of funding paid by the Commonwealth when it issues AAA-rated government debt to finance these investments is substantially below the returns generated by the securitised portfolios, which are also mortgage-loss insured, taxpayers cream a near riskless profit.
We can run a rough calculation of the profits taxpayers are making by quickly quantifying the differential between, say, the cost of 3 year government debt and the returns yielded by the government's investments to date. My numbers suggest that this spread is about 1.3 per cent per annum. In fact, right now it looks a little higher than this, but let's round down.
So once the Commonwealth puts all of its $16 billion to work, which will happen relatively soon, taxpayers will in theory be left with $208 million every year after the Commonwealth has fully repaid its interest costs using the crude assumptions I've outlined above. And because the Commonwealth is investing in AAA-rated assets, there has been no increase in net government debt. Of course, the actual returns might be less than this, but the illustration gives you a guide as to how the numbers should broadly work.
This analysis is verified by referencing the Commonwealth's reported performance. In its 2008-09 Annual Report, the AOFM confirms that it has been able to invest in residential mortgage backed securities at spreads that were "attractive in historical terms". Prior to the GFC, these assets would sell for around 0.2 per cent to 0.3 per cent over the benchmark rate. The AOFM was able to buy them much more cheaply at prices that implied far higher returns of between 1.2 and 1.4 per cent over the benchmark.
I only actually looked at the annual report once I had written the first version of this note. My estimates were uncannily close. The AOFM reported that on its first $6.2 billion of investments it captured a "a weighted average margin of around 1.3 per cent per annum over the one month bank bill rate...The average credit margin of around 1.3 per cent per annum on the RMBS portfolio is above the AOFM’s cost of short-term funding, which has historically been below the bank bill rate." The AOFM is telling us that the spread they earned for taxpayers was actually greater than the 1.3 per cent I had assumed.
I should add here that the Treasury was never supportive of this policy measure. Bearing this in mind, it is interesting to see that the AOFM emphasises that during the GFC it could have acquired these home loans for even cheaper fire-sale prices (forgetting that the spreads at which they bought the assets for were already 10x higher than the long-term pre-GFC averages).
The AOFM notes that this would have in theory further boosted its margin by another 0.8 per cent (or 80 basis points) to 2.1 per cent per annum in total over the benchmark rate. Yet the AOFM comments that if the "investments had been priced at yields 80 basis points wider, they would not have provided a viable source of funding for housing." And the fact is that these fire-sale prices no longer exist, in large part because of the Commonwealth's successful effort in restoring liquidity to the marketplace.
I guess to hedge its policy bets, the AOFM discloses that if it were to value its investments using the fire-sale GFC prices, it would have technically bought the assets at a "mark-to-market" loss. This is an irrelevancy for four reasons: first, there was no market at the time for the assets, which is why regulators around the world stopped using the mark-to-market approach since the prices were bogus (nobody, including the RBA, thought that a spread of 2.1 per cent was remotely near fair-value); second, the returns the AOFM cares about are the actual returns generated by these investments, which are determined by the positive 1.3 per cent spread the AOFM reports; third, the AOFM's investments were never motivated simply to produce the highest possible returns--the whole point of the initiative was to supply liquidity to a market that was shut; and finally, if one did want to value the AOFM's portfolio today using mark-to-market prices it would show a positive return given spreads have now fallen back below the AOFM's entry levels.
In addition to the obvious economic benefits, this initiative has been vital to maintaining competition in Australia's banking and finance sectors. The first $8 billion of AOFM investments quite literally prevented the failure of many smaller lenders during the GFC.
It could nevertheless be improved--I have long argued in favour of a new APRA-managed licencing regime for securitisers that dictates minimum solvency, documentation, structuring, and information disclosure standards. Self-regulation will not work. The industry needs to understand that there are no guarantees that the AOFM will be there in the next crisis. Policymakers should also impose a regulatory cost for the taxpayer assistance (in addition to the pure economic cost tied to the funding), especially considering that securitisation is an 'over-the-counter' market that has not ever been subject to formal oversight.
Finally, I would like to see the government develop a much longer-term, strategic policy framework for supporting liquidity in the non-intermediated securitised housing finance market, just as it does in the bank intermediated sector. It could very easily do this in concert with the imposition of the abovementioned APRA licencing regime. The AOFM is currently just implementing a policy that was announced on the fly in the midst of a global meltdown. At some point, the government has to step back and resolve a more thoughtful long-term plan. This should be a combination of new regulation in exchange for the provision of AOFM-style liquidity support during future emergencies.
