How do the lenders fund mortgages?

Andy836 said:

SVR mortgage holders are compensating for those on trackers and to a slightly lesser extent for those in default.

Click to expand...

Don't think so.

BOI's NIM on its tracker book at the end of H1 2015 was 23bps. The NIM on the tracker book is certainly much lower than BOI's overall NIM ((2.21%) but it's not true to say that (performing) trackers are loss making or need to be compensated by other loan products.

Bear in mind also that mortgages account for just over half of all loans - most unsecured lending would be on a much higher margin.

On the other hand, 7% of BOI's owner occupier mortgage book and 16% of BTL mortgages were in some form of forbearance or arrears at the end of H1 2015.

@Sarenco,
Its is very easy to say that currently performing trackers are loss making.
No way would that 23bps gross margin cover the associated admin costs, fixed asset overheads, etc.
Now, on the face of it, your comments on other unsecured lending is valid however, BOI provides segmental information which can be analysed. This includes its ROI Retail Book.

So lets look at BOI's ROI Retail figures for H1 2015,

ROI Retail Assets
BOI ROI retail had $26.8bn in assets.
This is split $26.8bn mortgages (resi & BTL) and $1.5bn other credit products.
This can also be split $13.8bn tracker, $13bn SVR.
So only 5.6% of ROI Retail assets are "Other unsecured lending which would attract higher interest rates".

Analysis of Net Interest Income
ROI Retail net interest income was $530mm (annualized $1,060mm),
Apply that 23bps NIM from its trackers to the $13.8bn trackers and you get $15.9m (annualized $31.8mm),
That means $514mm in H1 2015 ROI Retail net interest income came from non-tracker mortgages.
This needs highlighting -
Non-Tracker - 97% of all BOI ROI Retail Net Interest Income comes from 48% of the Book.
Tracker - 3% of all BOI ROI Retail Net Interest Income comes from 52% of its Book

Analysis of ROI OpEx
ROI OpEx costs were $394mm
We need some assumptions here:
ROI Retail Total income was $731mm which includes the $530mm (73% of total income) net interest income and $201mm (27% of total income) in Other Income (personal account fees, transaction fees etc)
Now, we know Other Income is capital light but is more labor intensive so lets invert the related costs and assume 73% of OpEx related to "Other Income" and 27% relates to Net Interest Income.
27% of $394mm is $106mm would reflect an 80bps admin rate based on the $26.8bn in assets. This is not unacceptable.
So, lets allocate this cost evenly across the loan books = $52mm to SVRs and $55mm to Trackers.

Depreciation
This was $22mm so allocate this as follows 20% trackers, 20% SVRs and 60% Other
This equals $4.4mm Tracker, $4.4mm non tracker, $13mm Other

Tracker Pre Impairment Profits
Tracker Net Interest Income $15.9mm
Tracker OpEx ($55mm)
Tracker Depreciation ($4.4mm)
Pre-impairment Op Income/(Loss) = ($43.5mm)

Next we look at the ROE
Total BOI Equity was $10.3bn
The ROI Retail book represented 20% of Total Assets so lets allocate 20% of capital to the ROI Retail Book which equals $2.1bn
On a straight up basis, ROI retail generates approx $513mm in annualized net income (after assumed 12.5% effective tax rate) which provides for a healthy 21% return on Equity.
However, if we exclude (i) 52% of the equity and (ii) the above Tracker NIM and assumed associated costs we'd have $715mm in Total Op Income and $262mm in net income (annualized $524mm) over an equity balance of $1bn we'd have a 52% ROE.


Right, so yes, BOI trackers wash their face on a gross margin basis, but they are not profitable on an operating basis and cannot provide any excess income to cover impairments and are wasting valuable capital.
The return on equity, majority of the opex and funds to cover impairments are all being funded by those on SVRs.

Andy836 said:

@Sarenco,
The return on equity, majority of the opex and funds to cover impairments are all being funded by those on SVRs.

Click to expand...

Absolutely agree with that statement.

It is perfectly clear that the vast majority of BOI's return on equity, opex and impairment costs are being funded by the return on loan products (secured and unsecured) other than tracker mortgages. I'm not trying to suggest otherwise.

However, you stated that "SVR holders are compensating for those on trackers and to a slightly lesser extent for those in default". That would suggest that trackers are loss making and need to be compensated for by other loan products. That's simply not the case.

The cost of administering a back book of performing mortgages is trivial. Really, what substantive administration is there? The process is almost entirely automated.

Bottom line - if performing trackers were losing BOI money, wouldn't you think they would be offering some incentives to borrowers to pay them off early?

Sarenco said:

Absolutely agree with that statement.

It is perfectly clear that the vast majority of BOI's return on equity, opex and impairment costs are being funded by the return on loan products (secured and unsecured) other than tracker mortgages. I'm not trying to suggest otherwise.

However, you stated that "SVR holders are compensating for those on trackers and to a slightly lesser extent for those in default". That would suggest that trackers are loss making and need to be compensated for by other loan products. That's simply not the case.

The cost of administering a back book of performing mortgages is trivial. Really, what substantive administration is there? The process is almost entirely automated.

Bottom line - if performing trackers were losing BOI money, wouldn't you think they would be offering some incentives to borrowers to pay them off early?

Click to expand...

I outlined the costs above and the fact that on an operating basis trackers are loss making. If you dispute the OpEx costs & my assumption outlined above please highlight which ones and why.
Bank Opex costs are real and have to be covered from existing income. Trackers do not provide enough income to cover the cost of their administration, their funding and their provisions.

The cost of administering a Back Book is not trivial (i cite 80bps above, I'm open to objections). There are tons of costs involved, which do have to be covered. Compliance, legal, statements, funding pension deficits, interest certificates, associated deposits. Just think about it, there's $27bn in assets in the retail book, there has to be $27bn in liabilities to cover those on the balance sheet. Are you saying the Bank can generate those liabilities for free? There's no admin cost in generating short term retail deposits, rolling them over every 1, 2 or 3 months, dealing with people breaking out of fixed term deposits, IT costs?

If what you said was true Bank OpEx would be on the floor at present as they're doing very little new business. Whether people like it or not, the back books have to cover the costs of generating "new business" as without "new business" the company is in wind down and will not survive.

Why aren't BOI offering incentives to have them paid off? They are 25 year products, at this point in the cycle they're loss making, doesn't mean they'll be loss making in 2 years or 4 years.

I believe the main focus of elevated SVR rates is incorrectly placed on existing default rates. This is clearly a big problem and a big costs but the impact of trackers is hugely under represented in the media commentary. They should be shoved into Nama, put on long term repo with the ECB or securitized.

Again, I would emphasise that I am only talking about the performing tracker book.

What legal or regulatory costs are associated with a performing book of trackers?

What substantive administration costs are associated with a performing book of trackers? 80bps - not a chance! Are you extrapolating administration costs from a securitised mortgage book? Don't forget that securitised books will include loans that will or could go bad in the future.

What does provisioning have to do with Opex?

What is going to happen in 2 or 4 years time to transform this into a profitable book? Are you assuming that BoI's OpEx or cost of funds will further reduce over this period?

I am of the firm opinion that your core conclusion is entirely misplaced.

If trackers were loss making for BOI, then logically they would be offered for sale (or redemption) at a discount. The fact is that BOI's cost of funds is now low enough that the performing tracker book is modestly profitable.

@Sarenco
I've time to look at this tomorrow.

I specifically split impairments from Opex when I analyzed their ROI retail accounts above.

Your argument seems to be based on the assmuption that trackers
- Have zero servicing costs,
- Have no requirement within the spread to cover potential losses,
- Have no, non interest, cost of funding,
- Have no cost of capital,

Your point about "if trackers are not profitable, why doesn't BOI sell them" is illogical. Well, 3 years ago they were definitely not profitable yet BOI didn't sell them or offer redemption incentives.

Why would they sell them? Who'd buy them and at what price would they buy them? Currently, they are making sufficent margin elsewhere to compensate i.e. there's no reason to take the hit on selling them when they can make SVRs cover the cost.

Sorry Andy but I really don't have time to deal with this today.

I have already agreed with you that the majority of operating expenses, impairment costs and capital costs are being funded by loan products other than trackers but it does not follow that performing trackers are loss making at BOI's current cost of funds.

I think the disagreement is over the definition of "loss-making", is it not? Your definition of loss-making only includes after its own costs of borrowing, period. But why?

If trackers can only be made profitable when the bank has to rely on that "the majority of operating expenses, impairment costs and capital costs are being funded by loan products other than trackers" then the commonsensical conclusion is that trackers are not loss-making only as long as the bank is able to extract additional funds from the SVR to compensate for "the majority of operating expenses, impairment costs and capital costs", therefore, as Andy836 said, "SVR mortgage holders are compensating for those on trackers". In other words, if 100% of loan book were trackers, it would have been loss-making as the bank would not have been able to compensate. Ergo, SVR are high (also) because of inability to extract from trackers what the banks would have normally been expected to do for their loans.

Andy836 said:

This is completely incorrect.

Click to expand...

Andy836, I of course meant that SVR/fixed rates of lending did not reflect the ECB rate. I obviously do not dispute that tracker mortages reflect those.

But I and many other good people of Ireland obviously do not have these magic beans and our rates do not reflect the ECB rates. But tracker or no tracker, they did up to 2008, even from late 2008 to mid 2010 or so the margin between the ecb rates and SVR rates never exceeded 1.5-1.6% or so. So it was reasonable to expect until mid 2010 that all lending rates reflect the ecb base rate more or less.

And Merry Christmas everyone!

If your point is that a significant part of the reason that SVRs are high is because lenders cannot pass on elevated operating, impairment or capital costs to borrowers on trackers, then we are in violent agreement. However, that's not the same thing as saying that performing trackers are loss making.

I'm really not sure how your expectations in 2008, reasonable or otherwise, assist your position.