Transcript of Question & Answer Session Mortgage Insights From Securitisation Data

Marion, we have some questions coming from the audience, if you have a few minutes.

Yeah.

First one should be a reasonably quick answer. How many users of the data are there, other than the RBA?

That is a question which I cannot answer. I’m not sure whether the ASF knows, but they provide the access for other users. We have access to different data than what is broadly available. There are some parts of the data with privacy concerns, there are self-securitisations as well. We’re the only ones who are accepting that as collateral. The issuers have been asked to provide statistics on those, but I can’t actually tell you how many users there are. I’m not the right person to ask. I don’t own the data.

Fair enough, next question, the last graph you showed, new borrowers pay lower interest rates than existing ones. You mentioned this different composition for those two segments of the market. What measures could help improve this lack of composition presumably, in the existing model, I suppose?

I wanted to take you back, so this is the last graph which shows a large gap in there, so there’s a discount. That does point to different competition. It’s not a recent feature. Competition in any market, including the financial market, is a complex issue. A number of agencies have been tasked to look at that. Most recently, the productivity commission has been asked to conduct an inquiry into competition in the financial system. They have asked for submissions, which I believe are publicly available. Ours certainly is.

One barrier to competition, amongst many, and the one which we have often pointed out, is lack of transparency or the availability of information. For example, in the example here, borrowers could refinance if they knew where they could get the lower interest rates and just change their loan. The RBA has highlighted more transparency as one of the steps that could be taken in that space and availability of data, in our view, is an important ingredient to transparency. The example here is the securitisation data has allowed us insights into the structure of mortgage interest rates that we didn’t have before.

Our next audience question, there are a few coming through now, what trends are you seeing in delinquency and hardship?

I think I have to pass on that one because that is a question for the financial stability area. I don’t tend to focus on this. If you want to know what we know, the Financial Stability Review is a great publication on that. The securitisation dataset is also not necessarily the best one to look at. As we know, there are certain conditions to securitise a loan. Short of having the population, it’s been unclear whether it’s representative in that respect or not.

That links quite well to the next question then. Does the much lower rates of investor and interest only loans in marketed RMBS, imply banks balance sheets carry a materially weaker credit quality than the securitisation product they place in capital markets?

I think that is for the relevant banks to argue. Of course, the loan itself isn’t the risk of the security. You’ve got tranching happening. You’ve got subordination happening. We have some of those in the books. We take only the AAA tranche. I think it’s probably for each individual investor to decide which path they take. I would say I can’t really speak to that because I don’t think the two things are comparable.

A question on the comparison between principal repayment in interest only versus principal and interest loans, asking does the dataset show that the amounts of principal paid off over five year terms is roughly the same between interest only and principal and interest loans? The suggestion being made is that interest only borrowers are simply paying it off in less frequent and larger chunks, rather than in contracted fortnightly repayments?

That’s a very good question. Thank you for the research agenda. I’ll be going back and looking into that. We haven’t looked at this specific question. I think one of the issues is also we have only two years of data so far so that stretches you just that far. There’s a little bit of historic data coming in. I haven’t looked at that specific question but watch this space.

You can give us the answer to that when you come back next year I hope.

We’ll see. If I like it.

Maybe just time for one more. How does the RBA treat balances in offset accounts when analysing LVRs?

In this specific exercise, we have excluded the balances in offset accounts, because they can be redrawn, so it’s a question of how sustainable any position is. There are probably valuable exercises where you want to include them. I know that there are a couple of buffer graphs floating around in our financial stability areas which come from thay dataset. So we’re looking at that for different questions. But for the loan to valuation ratio here, we have actually excluded them.

Thanks Marion. There are a couple of questions we might not have time to get to. Hopefully people will follow up with you on email if they want something else.

Absolutely. There’s Cécile here. I’m sure she will also point you to the right person to answer the questions.