It is pretty ironic that post-GFC bank regulation is focusing on the banks’ capital when their fast-growing attackers, the so-called peer-to-peer lenders, have no capital at all.
It means the next downturn, whether crash or recession or both, will be a very big turning point for the banking industry.
Either you need a capital buffer for safe saving and lending, or you don’t. If we find out that you don’t, then -- to get technical about it -- the banks are buggered.
The global P2P lending book is still less than 0.1 per cent of bank assets, so a long way to go yet, but growth is more than 100 per cent per annum. It would be more, I’d suggest, if they had been through a recession already.
The contest between banks and this new wave of non-banks comes down to a different approach to managing risk in unsecured lending.
Banks wear the losses themselves and use capital as a buffer. The lenders (depositors) only lose if the bank’s capital runs out and/or there’s a run.
In the GFC, many of them found their capital wasn’t enough, or their depositors believed that to be true, so, there was a run, and they went broke (although that was largely about secured lending on housing -- their second buffer, the loan to value ratio (LVR), evaporated as well because they lent too much and values were overstated).
There are two types of risk management in unsecured P2P lending:
1. The sort pioneered by Zopa, The Lending Club and Society One, which involves each borrower getting a risk rating and paying an interest rate that’s set accordingly. The lenders are expected to manage their own risk by having a diversified portfolio of loans. In a way, it’s like DIY CDOs -- collateralised debt obligations -- which were the packages of secured debt that were popular pre-2007 but blew up because the LVRs evaporated when housing collapsed.
2. The model developed by RateSetter in the UK which involves each borrower paying a “Risk Assurance Charge” into a “Provision Fund” which can be drawn on in the event of default. In effect, it’s an insurance model. Borrowers are still charged different interest according to risk assessments, but lenders have some extra protection.
The two models have divided P2P lending into wholesale and retail.
The first model really only suits wholesale lenders -- hedge funds and super funds -- because they have large enough books to spread and manage their risk. As a result, operators like Lending Club have basically become an alternative asset allocation for managed funds and calling them ‘P2P’ is a bit of a misnomer.
RateSetter, on the other hand, gets funds almost entirely from retail investors: in Australia, where it has only recently launched, its average investment is about $2,000, although a lot of people have so far put in less than $100 to try it out.
The key to success for both types is the minimal interest rate spread between lender and borrower: lenders get a higher yield and borrowers pay lower interest.
The firms operate lean, with algorithms doing a lot of the credit risk assessment, but, most importantly, they have no capital to service. Apart from RateSetter’s Risk Assurance Charge, it’s DIY capital.
As a result, a lender can get a yield of 8-10 per cent for 3-5 years with borrowers still paying much less for an unsecured loan than they would at the bank.
P2P lenders typically charge a fixed, transparent fee, usually around 10 per cent, rather than the spread between deposit and loan which is both undisclosed and whatever the bank can get away with. Most banks are also colossal, expensive bureaucracies, working hard to get their expense ratios down.
P2P lenders are sometimes called internet or technology businesses, but that’s not true. They are simply another form of financial intermediary, in the way that Uber is another form of taxi service and AirBnB is another form of accommodation provider -- it’s just that the internet makes it easier to co-ordinate the providers.
Meanwhile, banks are under pressure to hold more and more capital. Today, for example, the CEO of the Myer Family Office and former head of institutional banking at ANZ, Peter Hodgson, told The Australian he thought banks needed to hold more capital, partly because their LVRs are too tight.
The discussion has been given extra bite by this week’s comments from Treasury Secretary John Fraser that Sydney real estate, and parts of Melbourne, were in a “bubble”.
When asked about what risks that represented for the banking system, Fraser replied: “I am very, very confident that APRA is addressing this problem properly.”
That is: making them hold more capital and restraining the growth of their loan books to less than 10 per cent a year.
The P2P folk, meanwhile, are happily growing their books at more than 100 per cent a year -- all unsecured, no capital. At this stage they are basically going after the car loan market, but presumably they will start accepting real estate as security at some point. Now that could be disruptive.
But this new ‘co-ordinated’ system of banking won’t be tested until there’s another spike in defaults.
Although the first P2P lender -- Zopa -- started in 2005 and Lending Club a year later, when the last crash and recession struck in 2008, they weren’t anywhere near critical mass and were barely even discussed.
In those days, the non-bank sector consisted of investment bank securitisers that were packaging secured and unsecured loans and selling them to yield-hungry investors for a margin.
They were found out in 2008 and basically disappeared; securitised mortgages have only now just started to recover as an asset class.
The unresolved question about the unsecured P2P operators is: how will their lenders fare in the next recession? Will they be better or worse off than those who left their money in the intermediaries with capital (banks)?
The P2P lenders I’ve spoken to say: “Bring it on”. They know their business model won’t be fully trusted until they have been tested by a recession … and have passed, of course.
And then, look out.