Good work, Snoopy, but all a bit redundant now that UDC (ANZ) has announced that it intends to repay all outstanding debentures later this year in favour of "alternative financing". Probably means hat ANZ will increase its support.
Printable View
Good work, Snoopy, but all a bit redundant now that UDC (ANZ) has announced that it intends to repay all outstanding debentures later this year in favour of "alternative financing". Probably means hat ANZ will increase its support.
Thanks for the update Macduffy. I hadn't seen the 16th Jan 2019 press release you refer to and so went straight to the UDC website.. They have certainly closed the shop door to new customer debentures and given notice that existing customers with will have their debentures redeemed. The door is still a tiny bit open though (my bold):
"If a final decision is made to repay all existing Secured Investments, ..."
And it seems the call account survives.....for now. I presume the 'call account' is secured?
I still think it is worth following UDC though, because they are probably the closest finance company to forum favourite 'Heartland' and it is always useful to have a measuring stick. But if no public money is to be taken in, then UDC becomes a fully contained subsidiary of the ANZ bank. So I guess public reporting will stop, and they will sack the board,
SNOOPY
You are correct in stating UDC is a useful measuring stick for HGH.
I hope ANZ taking it in house does not weaken it.
My thoughts are mixed.I would have liked it to merge with HGH or retained [semi] independant from ANZ.
Being listed would have worked.I am sure NZders would have continued to fund it,but without the ANZ support it was always going to be difficult finding "fair" or "true" value for it.
UDC Heartland EBT Loan Book EBT/Loan Book EBT Loan Book EBT/Loan Book FY2009 $34.024m $1,829.156m 1.86% FY2010 $45.012m $1,968.771m 2.29% FY2011 $46.382m $1,948.552m 2.38% FY2012 $58.476m $2,014.473m 2.90% $29.377m $2,078.276m 1.41% FY2013 $66.787m $2,065.117m 3.23% $36.540m $2,010.376m 1.82% FY2014 $83.501m $2,272.081m 3.68% $57.416m $2,607.393m 2.20% FY2015 $89.750m $2,347.163m 3.82% $76.304m $2,862.070m 2.67% FY2016 $88.835m $2,573.030m 3.45% $87.689m $3,113.957m 2.82% FY2017 $91.639m $2,911.514m 3.15% $99.568m $3,545.897m 2.81% FY2018 $101.664m $3,222.430m 3.15% $116.361m +($1.3-$4.8-$0.6)m $3,984.941m 2.82%
Note:
1/ UDC data for FY2018 is drawn from the 'UDC Finance Annual Report 2018' 'Statement of Comprehensive Income' (EBT) and 'Balance Sheet' (Loan Book Balance).
2/ Heartland Bank data for FY2018 'Statement of Comprehensive Income' (EBT) and 'Statement of Financial Position' (Loan Book Balance).
3/ All EBT figures are before 'credit and impairment charge'.
4/ FY2018 Heartland result adjusted to reflect the before tax effect of a $4.8m gain from a formerly written off property now sold, a $0.6m gain from the sale of the bank invoice finance business and $1.3m in one off adjustment for internal system reintergation costs.
Note that the absolute figures year to year are not comparable between UDC and Heartland. This is because Heartland has a physical branch structure whereas UDC works out of ANZ bank branches. The underlying cost structures of both protagonists are not the same. Furthermore Heartland's balance sheet income earning base is more than just the receivables. The Heartland banking group has investment assets, there is goodwill on the books from high earning acquisitions, there are vehicles owned in lease out deals and rent from investment properties. That sums to 10% more income earning assets than just the receivables. Yet the increasingly higher earnings rate at Heartland now belies the higher cost structure that Heartland must have - impressive for Heartland?
The individual company year on year trend is interesting though. The EBT Margin for UDC continues to decline, even as the Heartland EBT margin continues to increase. After many years they are both converging towards the same value! Is the unusually large increase in the loan book size ('growth any any cost' to pump up a potential company sale price?) at UDC compromising the profitability for UDC going forwards? Given profitability for FY2018 has remained at previous year's lower levels, that could suggest the answer is 'yes'!
SNOOPY
Time to normalise the UDC figures for 2018 so they can be compared more directly with the likes of Heartland Bank. Heartland in FY2018 had selling and administration expenses of $80.433m (Heartland FY2018 report 'Selling & Administration Expenses', note 5). UDC had total operating expenses of $34.838m (UDC Financial Statement 2018, note 4). That is a difference of $45.595m. The two are comparable in that they have a similarly sized loan book (UDC:$3,222.430m, Heartland $3,984.941m). If we add the operating cost difference figure onto the UDC cost structure, what would that do to the UDC operating margin (EBIT where 'I' is the credit facility interest only) on assets?
FY2018: ($160.401 - $45.595 + $6.592) / $3,222.430 = 3.77%
Note: UDC do not have a branch network of their own, but operate through ANZ bank branches in New Zealand. The $6.592m added back represents the adding back of 'fees paid to related parties' (ANZ). These are part of the $45.595m 'extra operating expenses' (calculated above, using figures from Financial Statements 2018 note 4). The $6.592m could be thought of as a contribution to the ANZ branch network that allows UDC to carry on business as normal. But what I am interested in is the difference in operating cost of a finance company with and without a branch network. So this $6.592m which largely reflects a 'branch network allowance payment' must be removed from my comparison.
For Heartland over FY2018 the equivalent calculation of EBIT (where 'I' represents interest paid not connected to debenture holders, and E represents the earnings before write downs) is as follows:
($116.361m+($1.3-$4.8-$0.6)m+$25.380m) / $4,397.350m = 3.13%
This recent year trend in the underlying margin at UDC and Heartland is compared below:
UDC underlying EBIT Margin Heartland underlying EBIT Margin FY2018 3.77% 3.13% FY2017 3.44% 3.18% FY2016 3.07% 3.47% FY2015 3.53% 3.57% FY2014 3.37% 2.66% FY2013 2.58% 2.32%
(Note: I have recalculated the figures from Heartland in relation to past years. This is because I have changed my mind about what Heartland assets on the balance sheet are 'income earning', and produce those Heartland earnings)
In this context, FY2016 looks like a negative aberration for UDC. But the EBIT margin for Heartland is showing a declining pattern over the last four years. When UDC was put up for sale in FY2017, could the earnings figures for UDC have been manipulated upwards over FY2017 in what was in reality a softer market? Just to make UDC's sale prospects look better? 'Earnings' in this context is a special kind of EBIT in which I have eliminated impairment charges. So I have removed the ability to reclassify assets as 'problem assets' in the current year. Although I should note that wouldn't stop UDC taking on more assets (loans) as a whole,- and they might include more problem assets. Even if they were not classified in that way - yet! That will only come out by inspecting the asset provisioning in subsequent years. Yet the ability to manipulate UDC's re-balancing of bank loan to debenture structure remains as an influence this EBIT result. As debenture loans from people on the street are replaced by ANZ in house bank loans, then EBIT goes higher. Because extra interest charges due to ANZ replace what was formerly an 'in house cost'. It is therefore the increasing mismatch between 'in house loans' and 'in house money used to support those loans' (because this implies more bank funding to bridge the gap) that is driving the underlying EBIT for UDC higher in 2017 and 2018.
It strikes me that what I have just shown here is that if you can disregard your external borrowing costs, then your underlying earnings will go up. Well, duh, yeah! Such an edict is so obvious it doesn't need proving. So maybe this post is just a waste of time? Sorry you had to read it! Subsequent to starting this post, I have found out that UDC owner ANZ is paying out all UDC debenture holders this year. As this will increase EBIT, perhaps the moral of the story is that those finance companies that take deposits from the public as less valuable than those that simply use parent bank financing. OTOH such finance companies can only exist on the whim of a parent funding bank. So perhaps the moral is: "Do not use EBIT as a valuation multiple for finance companies!"
SNOOPY
HLB (FY2018) UDC (FY2018) Agriculture Forestry & Fishing: $808.452m (18.9%) $573.893m (17.5%) Mining: $19.222m (0.4%) $19.976m (0.6%) Manufacturing: $70.822m (1.6%) $60.655m (1.8%) Finance & Insurance: $337.241m (7.7%) $66.999m (2.0%) Retail & Wholesale Trade: $238.064m (5.4%) $419.054m (12.8%) Households: $2,105.231m (48.0%) $965.008m (29.4%) Property & Business Services $399..973m (9.1%) $183.128m (5.6%) Transport & Storage: $206.592m (4.7%) $437.710m (13.3%) Other Services: $184.826m (4.2%) $558.206m (17.0%) Total $4,390.423m (100%) $3,293.630m (100%)
Note:
1/ Heartland loans pre impaired asset adjustment. UDC loans post impaired asset adjustment.
The Heartland loan book has grown by 11.6% over FY2018 (ended 30-06-2018), compared to a 10.7% growth over at the UDC loan book over the nearest equivalent period (FY2018 ended 30-09-2018).
Discussion UDC
At UDC, the press release of full year results highlights were:
1/ Motor vehicle lending increasing by $217 million (+18% to $1,200m),
2/ Commercial lending growing by $50 million (+4% to $1,300m) and
3/ Equipment dealer lending was up $12 million (+6% to $200m).
These were the same three category increases that UDC chose to highlight in FY2017, even if the motor vehicle increase was much more modest in percentage terms than FY2017. Curiously those categories do not equate to the category loan disclosures made in the UDC annual report. Yes 'Household Lending' was up by $144m. But Transport and Storage was actually down (again), this time by $5m. So once again, I would guess that most of the increase in motor vehicle loans were made via retail sales at car dealers. The largest dollar increase in any category was for 'Households' , the $144m gain being up 17.6%. 'Household receivables' include private car sales. UDC noted the automotive sector was 'slowing down' and business confidence was lower, a striking juxtaposition to UDC's own experience of motor vehicle loans growing strongly:
1/ A sign of UDC's growing market share in motor vehicle finance?
2/ Or a comment on a slow down in business in the second half of CY2018?
In FY2018 UDC make a big deal of 'construction lending', while Heartland doesn't even give construction a lending category. Construction lending at UDC increased to $451.173m (+10.3%) up from $408.967m. Yet a similar sized increase of $52m (much greater in percentage terms, +14%) happened in 'retail and wholesale'. And this was not mentioned in UDC's press release of results commentary.
Discussion Heartland
Meanwhile at Heartland, on-line lending is highlighted as growing fastest. Heartland's share of the Harmony platform was up 61% to $152m, albeit off a small base. Whether this is indicative of the wider incremental uptake of digital loans via other Heartland digital platforms I am not sure. The much trumpeted wider digital platform achievements of:
1/ helping more kiwis into franchises and
2/ to purchase plant and machinery,
are all subsumed inside wider lending categories. Yet Heartland did report small business lending via the digital base up from $45.4m to $89.9m: a 98% increase (but still only making up 2% of the total overall Heartland business). Of course that 98% increase may include some business loans that would have happened anyway, but were redirected from what would have been paper loan applications.
The proportion of Rural Loans decreased for the first time in five years, now 18.9% of the total - down from 21.3% in FY2017. This is useful in de-risking the Heartland loan book going forwards. And it is consistent with Heartland's announced strategy of moving away from 'larger business and rural relationship lending', to shorter term seasonal lending, like livestock finance. This kind of targeted rural lending will remain a cornerstone of Heartland's business going forwards, along with reverse mortgages (+23% to $1,130m), motor vehicle loans (+16% to $955m) and small business loans (+7% to $1,066m). In contrast to UDC, Heartland see further growth in motor vehicle loans. Heartland also see no impact on their Australian reverse mortgage business from the Australian Federal government starting a domestic reverse mortgage scheme for income supplementation. Yet 20% of existing Heartland reverse mortgages are used to generate extra income! I do hope for shareholders sakes that Heartland are not 'seeing only what they want to see'.
Heartland has, post results, restructured their business to allow the reverse mortgage business in Australia to keep growing outside of its' Heartland Bank company cousin in New Zealand. 'Household Loans' as a whole category grew by 23%, matching the growth of reverse mortgages that now make up more than 50% of 'household loans' category. That is important, as it shows that reverse mortgage growth is being matched by other household loan growth, This largest category growth rate is ahead of what UDC are achieving. UDC grew their household loan book by a still impressive 17% YOY without offering reverse mortgages. The worst performing loan category for Heartland was 'manufacturing' with gross dollar loans falling by 7.5%, compared to a small increase at UDC. In both cases we are talking about less than 2% of all receivables - not material to either companies' annual result. Evidence that New Zealand is progressing further down the path of a 'post manufacturing' era?
Risk Concentration Assessment
My overall impression of comparing these two protagonists are that they are becoming more alike. The stand out differences of UDC highlighting $450m worth of 'construction lending' and having $200m more invested in both the 'transport and storage' and 'retail and wholesale', while Heartland has a fast growing platform of $1.1billion in equity release loans, notwithstanding. The greatest risk aggregation for both protagonists is in the 'household'/'personal and other services' category. But these loans are geographically diverse and cover a multitude of needs. The largest sub category in this is 'Australian Reverse Mortgages' at Heartland, which totalled $NZ677m at balance date. That is just over 15% of Heartland's loan book. But those loans are spread all over Australia, the largest concentrations being 25% in Sydney and 18% in Melbourne. The Sydney HER market represents just 4% of Heartland's total loan book, not enough to cause concern. My main area of concern at Heartland last year, of excessive lending on land to dairy farmers in particular, has been addressed.
SNOOPY
PGW has the largest % exposure to rural.
Since 2017 HGH rural exposure has reduced from $679.1 mil to $660.5 mil in 2018,while their overall loan book has grown .
The average rural loan size has been reducing from $252,500 in 2013 to $174,400 in 2018.
The term of rural loans has reduced from 53 months in 2013, to 43 months in 2018.
A very different picture from the one you paint.
Since PGW is a rural servicing company, that statement is undoubtedly true. But what about rural financing? PGW would claim that they don't do that any more, having sold off PGG Wrightson finance to Heartland a few years ago. But we PGW shareholders know that the new 'Go Lamb' and 'Go Beef' finance schemes for livestock is PGW Finance recreated in all but name. The loan balance of 'Go Product' was $39.4m in aggregate at balance date. Compared to the PGW asset base of $537.1m, this is 7.3% of all assets.
That means in terms of farm lending, Heartland is still far more exposed than PGG Wrightson, by more than a factor of two.
I did put a note in my post that it was a 'work in progress' when you read it Percy. But obviously it wasn't prominent enough. The text you are referring to relates to the previous year. And as you rightly say, things have now turned the corner for Heartland and those difficult rural loans. My offending post has now been rewritten!Quote:
Since 2017 HGH rural exposure has reduced from $679.1 mil to $660.5 mil in 2018,while their overall loan book has grown .
The average rural loan size has been reducing from $252,500 in 2013 to $174,400 in 2018.
The term of rural loans has reduced from 53 months in 2013, to 43 months in 2018.
A very different picture from the one you paint.
SNOOPY
Updating the actual bad debt write offs in relation to the size of the loan book at the end of FY2019. Section 7 in the UDC 2018 Financial Statements is named "Provision for Credit Impairment". Below, bad debts actually written off are compared against the 'provision for loan impairment' stated on page 3, the 'Statement of Comprehensive Income'.
UDC Bad Debt Write Off (Note 7: Provision for Credit Impairment) New Annual Bad Debt Provision (Income Statement) FY2014 -$3,300m +$18.633m = $15.333m $11.733m FY2015 -$0.659m + $12.162m= $11.503m $10.427m FY2016 -$1.297m + $11.055m = $9.753m $7.418m FY2017 $2.860m + $7.698m = $10.558m $5.929m FY2018 $2.196m + $6.679m = $8.875m $10.885m
Actual write offs look to be in a range of $10m to $12m,
UDC Write Offs
Putting these 'actual write offs' as a percentage of the end of year loan book gives them better context Note that:
1/ the 'actual write offs' are found in the annual change of the holding provision for bad debts (note 7 'Provision for Credit Impairment' AR2018) and do not directly correspond to the top up expenses for this provision that may be found in each annual income statement.
2/ the denominator in the following calculation is the 'carrying value' of the Net Loans and Advances, This has already been adjusted for the provision for credit impairment, unearned income and deferred fee revenue and expenses.
FY2014: $15.133m/$2,272.081m = 0.666%
FY2015: $11.503m/$2,347.163m = 0.490%
FY2016: $9.753m/$2,573.030m = 0.379%
FY2017: $10.558m/$2,911.514m = 0.363%
FY2018: $8.875m/$3,222.430m = 0.275%
For FY2018 UDC the spectacularly low percentage of loan write offs continues.
Heartland Write Offs
For comparative purposes, it is informative to look 'over the fence' to Heartland Bank. See Note 6 ('Impaired Asset Expense' AR2018) to see the current year 'asset expense' calculation. Note that unlike UDC, Heartland writes off uncollectible debts or part debts directly from each annual profit result. In the calculation below, I have normalized these against the 'total finance receivables'. 'Total finance receivables' are already adjusted for any provision for impairment and the present value estimate of future losses (AR2018, Note 11 'Finance Receivables').
FY2012: $5.642m/ $2,078.3m = 0.271%
FY2013: $22.527m/ $2,010.4m = 1.12%
FY2014: $5.895m/ $2,607.4m = 0.226%
FY2015: $12.105m/ $2,862.1m = 0.423%
FY2016: $13.501m/ $3,114.0m = 0.434%
FY2017: $15.015m/ $3,546.0m = 0.423%
FY2018: $22.067m/ $3,984.9m = 0.554%
Summarizing and comparing the above information:
UDC Debt Write Off Heartland Debt Write Off FY2014 0.666% 0.226% FY2015 0.490% 0.423% FY2016 0.379% 0.434% FY2017 0.363% 0.423% FY2018 0.275% 0.554%
The question that rears its ugly head from the above data table is as follows:
Why is the impairment percentage so much lower for UDC in FY2017/FY2018 compared with UDC's past year results? Perhaps we had two really low impairment years? But if that was true, might we not expect a similar reduction in impaired loans over the same time period from the closely comparative Heartland?
A Change in Standards
IFRS 9 is a a new accounting standard, not yet adopted by UDC nor Heartland when their respective FY2018 accounts were published. IFRS 9 requires impairment losses to be provisioned for loans as they are taken out, based on ECL (Expected Credit Loss) rates. A retrospective loss due to already embedded loans, bumping up the 'impaired loan' balance, will be taken on the balance sheet of each company as follows (best estimate published);
Higher Impairment Aggregate EOFY Impairment Balance Implied Increase Increase in Deferred Tax Asset Net Effect on Balance Sheet UDC $11.4m $34.568m +33% $3.2m $8.2m Heartland $20m-$25m $32.495m +62%-77% $6m-$7m $14m-$18m
From this, it looks like UDC really do believe their loan book is only half as risky as that of Heartland.
Skullduggery or not?
UDC was put up for sale over the FY2017 financial year (before FY2017 accounts were published) but withdrawn from sale in FY2018 (after financial accounts were published). It would have been helpful to show the accounts in their best possible light leading up to any sale. So were the impairments at UDC for FY2017/FY2018 really that much lower? Or has some 'window dressing' gone on here? The sale of UDC has been put on indefinite ice, from 31st October 2018 (after balance date). So it will take until the FY2019 result at best for any 'window dressing', should it exist, at UDC to unwind.
SNOOPY
In my previous post on this thread, I have looked at 'impairment expenses' and 'impairment provision expenses'. But now I wish to turn my attention to include that period before an impaired loan becomes that way, and the accumulated 'impairment provision' itself. This means, I will look at loans that are vulnerable -those in danger of becoming impaired. In an ideal world, there might be no point in doing this exercise. We might expect vulnerable loans to come and go exactly like impaired loans, as the loan climate waxes and wanes. But in this non-ideal world I feel that we might learn something from looking deeper. Let's see....
UDC Vuknerable Loans
Finance companies have their own internal way of grading loans on their books. Within UDC note 10d (page 18 UDC Financial Statements for FY2018), lists the 'internal risk grading' of the all the loan assets on the balance sheet on a scale of 0 to 9. On this scale, 0 is the 'lowest risk' while 9 means a 'default'. I have added together loan classes 6 and above, a collection of loans for which I have coined the term 'vulnerable'.
UDC Vulnerable Loans Judgement Total Grade 6+ 2012 $975.744m +$80.745m +$55.403m $1,111.892m 2013 $1,157.111m +$83.790m +$24.814m $1,265.715m 2014 $811.700m +$92.366m +$34.883m $938.949m 2015 $904.338m +$81.156m +$32.640m $1,018.134m 2016 $1,127.677m +$96.727m +$17.657m $1,242.061m 2017 $1,201.747m +$133.791m +$11.618m $1,347.156m 2018 $1,221.379m +$111.290m +$16.780m $1,349.449m
The grade 6 and 'more risky' categories for EOY2018 added up represents a fraction of the total loans outstanding as follows:
$1,349.449m / $3,319.198m = 40.7% of total loan assets.
The offsetting accumulated 'impairment provision' on the books, not yet removed from the above total, is $34.568m (note 10d). This impairment provision represents:
$34.568m/$1,349.449m = 2.56% of UDC 'Vulnerable Assets'.
Heartland Vulnerable Loans
For comparative purposes it is interesting to see what happens when we derive the same statistics for Heartland bank. The situation is not strictly comparable, because Heartland has a different 'two box' credit risk system. The first box houses the so called 'Individual Behavioural Loans'. Behavioural loans consist of consumer and retail receivables, usually relating to the financing of a single asset.
There exists a second box of Heartland loans termed 'Judgement Loans' which are graded on the 1-9 system. Grade 1 represents a 'Very Strong' loan. Grade 9 represents a loan 'At Risk of Loss'. Grade 6 represents a loan that bank staff management should 'monitor'. A 'Judgement loan' within Heartland consists mainly of business and rural lending, including non-core property, where an ongoing and detailed working relationship has been developed.
The grade 6 and 'more risky' categories of 'Judgement Loans' plus the equivalently vulnerable 'Behavioural Loans' sum up to a total amount of what I define as Heartland 'Vulnerable Loans'.
Heartland Vulnerable Loans Behavioural Judgement Total Arrangement Non Performing Repossession Recovery Grade 6+ 2012 $13.750m $4.386m $2.740m $185.315m +$53.360m +$14.036m +$13.741m $287.118m 2013 $8.416m $2.226m $1.936m $198.370m +$18.034m +$21.518m +$27.761m $278.051m 2014 $7.571m $2.113m $2.113m $165.776m +$14.833m +$13.520m +$3.412m $159.338m 2015 $15.855m $3.087m $3.687m $99.849m +$14.937m +$4.514m +$7.082m $149.011m 2016 $14.923m $6.507m $7.171m $125.902m +$20.434m +$16.904m +$12.188m $204.029m 2017 $18.512m $4.956m $4.889m $166.155m +$27.669m +$16.749m +$2.556m $241.486m 2018 $46.728m $5.670m $5.490m $145.706m +$22.958m +$23.920m +$6.515m $256.987m
'Vulnerable Loans' for FY2018 represent a fraction of the total loans outstanding as follows:
$256.987m / $3,984.981m = 7.14% of total loan assets.
Impairment $29.671m (AR2018, Note 20a) has already been taken onto the book over the years. Add to this a reverse mortgage fair value adjustment of $2.824m. This total impairment of $32.495m represents
$32.495m / $256.987m = 12.6% of 'Vulnerable Loans'..
Comparing the Protagonists
A summarized comparative table between UDC (Year ending 30th September) and Heartland (Year ending 30th June) is below:
UDC Heartland Impaired Loans (A) Grade 6+ Loans [total Vulnerable](B) (A)/(B) Total Loans (C) (A)/(C) Impaired Loans (A) Total Vulnerable Loans (B) (A)/(B) Total Loans (C) (A)/(C) 2012 $38.481m $1,111.892m 3.46% $2,141,780m 1.79% $27.426m $287.118m 9.55% $2,105.702m 1.30% 2013 $37.460m $1,265.765m 2.95% $2,198,653m 1.70% $50.491m $278.051m 18.24% $2,060.867m 2.45% 2014 $31.805m $938,899m 3.38% $2,375.936m 1.34% $24.381m $159.338m 15.3% $2,651.754m 0.919% 2015 $31.529m $1,018,134m 3.10% $2,461.224m 1.28% $31.654m $149.011m 21.2% $2,893.724m 1.09% 2016 $28.909m $1,242.061m 2.33% $2,684.750m 1.08% $26.148m $204.029m 12.8% $3,140.106m 0.833% 2017 $29.278m $1,347.156m 2.17% $3,005.059m 0.974% $29.716m $241.486m 12.3% $3,575.613m 0.831% 2018 $34.568m $1,349.499m 2.56% $3,318.198m 1.04% $32.535m $256.987m 12.7% $4,017.436m 0.810%
Lot's of numbers here, so what does it all mean? I will start by defining the building blocks of my argument:
1/ 'Impaired Loans' is a judgement call on a portion of certain loans that have probably gone bad. 'Impaired Loans' is a provision in the accounts. When all hope of fully collecting such a loan is lost, then the appropriate 'impairment expense' is deducted from that provision and our financial institution moves on. 'Impaired Loans', which by definition are still on the books, must of necessity be measured by bank managers judgement. By contrast, an 'Impairment Expense' is a definitively measurable transaction.
2/ 'Vulnerable Loans' reflect those same bank managers judgments, even though most vulnerable loans are (as yet?) not impaired.. I feel it is fair to make a 'like with like' comparison between management's judgement at a transition level (Vulnerable Loans) with the same managers making a judgement of loans in a deeper level of distress (Impaired Loans).
The impaired loans as a percentage of total loans ( A/C in the table above ) appear similar for both protagonists, especially in more recent years. I would give a slight edge to Heartland here in having fewer impaired loans, although it might be a margin-of-error difference. However the subset of 'impaired loans' is a much smaller proportion of 'vulnerable loans' at UDC compared to Heartland. Could this be because UDC has a much smaller staff and so has to resort to a broader scattergun automated process in assessing how vulnerable their loans are? Whereas at Heartland there is more ongoing human contact with borrowers, so the loans that are vulnerable are easier to spot? If that is true of UDC, then marking more loans as vulnerable has not seen a comparative consummate reduction in the percentage of loans that become impaired. Is this a vindication of a more human interaction rich approach at Heartland? And if that is true will Heartland's move to more 'digital platform' lending see Heartland's percentage of impaired loans rise in the future? Perhaps the benefit of having a 'branch structure' throughout the country are not a 'last century' as some in Heartland management think?
In any event I feel that comparing numbers across columns is not the way to go. Differences across columns are just as likely to reflect inherent differences between two businesses, and not differences in performance. What we investors need to be looking at is consistency down the columns. And if consistency is not observed, then we need to figure out "Why not?" Looking down the A/B and A/C columns, I do see a trend of fewer impaired loans from vulnerable loans over the years. And fewer impaired loans out of all loans. That is a sign that both company's management are steadily getting better at doing their jobs. And that has to be good for both UDC debenture holders and Heartland shareholders.
SNOOPY
Time for my annual 'disentanglement' of ANZ.NZ from its UDC subsidiary. The ANZ.NZ is the largest bank in New Zealand. That means the way the bank behaves has significant implications for all investors in NZ, not just ANZ group shareholders and UDC debenture holders.
The information I need about the ANZ bank in New Zealand can be found here:
https://www.anz.co.nz/about-us/media...r-information/
UDC and ANZ New Zealand have the same balance date. So it is legitimate to work out the distribution of loans on their respective books using 30th September end of year data. First I need to:
1/ Slightly rearrange the ANZ (NZ) categories (ANZ September 30th 2018 Bank Disclosure Statement, p32) so that they link up to those listed in the UDC FY2018 Financial Statements. THEN
2/ I need to subtract the UDC equivalent figures (page 18, UDC FY2018 Financial Statements) to get the underlying ANZ bank figure.
(Note: Receivables for UDC in industry groups are listed after provisions for credit impairment are taken into account. OTOH, receivables for ANZ.NZ industry groups are listed before allowances for credit impairment are taken into account. This means the UDC figures are lower than they would be on a 'like for like' comparative figure basis. However the error is only 1.0% overall, not enough to undo the validity of this exercise in my judgement)
The results are below:
All ANZ.NZ = UDC + Underlying ANZ.NZ Agriculture forestry, fishing and mining: $20,936m (11.3%) $594m (18.1%) $20,342m (11.2%) Business and property services: $35,501m (19.2%) $183m (5.6%) $35,318m (19.5%) Construction: $3,092m (1.7%) $451m (13.7%) $2,641m (1.5%) Electricity Gas Water & Waste: $3,309m (1.8%) $15m (0.5%) $3,294m (1.8%) Finance and insurance: $19,324m (10.5%) $67m (2.0%) $19,257m (10.6%) Government and local authority: $12,868m (7.0%) $0.377m (0.0%) $12,868m (7.1%) Manufacturing: $4,764m (2.6%) $61m (1.9%) $4,703m (2,6%) Personal & Other lending: $75.796m (41.0%) $1,055m (32.1%) $74,741 (41.2%) Retail and Wholesale: $7,195m (3.9%) $419m (12.8%) $6,776m (3.7%) Transport and storage: $2,126m (1.1%) $438m (13.3%) $1,688m (0.9%) Total: $184,911m (100%) $3,293m (100%) $181,628m (100%)
As was the case last year, the loan allocation of ANZ.NZ with UDC removed, is little different the loan allocation of the whole of ANZ.NZ. This is no surprise. The whole of the UDC loan book is only 1.8% of the ANZ.NZ loan book. And ANZ.NZ itself (which you cannot invest in directly) is only a fraction of the whole ANZ operation, which is the ANZ vehicle listed on the NZX. However, the converse is not true.
UDC is very different from ANZ.NZ. In sector allocation percentage terms:
1/ the Agricultural exposure of UDC is 60% higher,
2/ 'Construction' and 'Transport and Storage' exposure are up by a multiple of 8, AND
3/ 'Retail and Wholesale' exposure are higher by a factor of 3.
The volatility of these three 'industry groupings' is testament to UDC being a much greater investment risk than any investment in ANZ itself.
The following inter-year table shows how UDC is funded by its 100% owner ANZ
UDC: Backing For Loans FY2014 FY2015 FY2016 FY2017 FY2018 UDC Shareholder Capital $341.412m (15.6%) $365.462m (14.6%) $423.247m (16.2%) $485.645m (16.7%) $550.944m (17.0%) ANZ Committed Credit Facility (Note 8) $280.000m (12.8%) $395.000m (15.8%) $595.000m (22.8%) $1,385,027m (47.6%) $1,762.003m (54.3%) Debenture Investments From Public (Note 8) $1,569.247m (71.6%) $1,736.026m (69.5%) $1,591.711m (61.0%) $1,039.133m (35.7%) $931,280m (28.7%)
There is a very significant change happening over the last two years, with the role of debenture holders in funding UDC much reduced as the ANZ parent seemingly looks to take over that role. It was confirmed in January 2019 that ANZ plans to pay out all UDC debenture holders over 2019.
SNOOPY
At the end of FY2015, UDC dined out on a glossy prospectus as the iconic provider of equipment across all New Zealand industry sectors. But shortly after this, the owner, ANZ Bank, decided to trim its corporate branches. UDC was one branch ear marked for the chop. The FY2018 UDC annual report was reduced to a comment less rough looking black and white e-document, designed to repel future investors, while still meeting legal requirements. I think it is fair comment to say that the failed sales process for UDC has become an embarrassment for the ANZ Bank. The solution seems to be to take UDC completely 'in house', stop the charade of UDC being seen as 'independent', while hoping the public forgets about the whole failed disposal, Yet this outcome was unknown if we wind the clock back to 2015. With pressure on the ANZ to improve its capital ratio, there was a real chance that ANZ could unload some of their risky account receivables by selling UDC. So how did ANZ plan to rejig the 'account receivable' balance so that the ANZ parent could get maximum benefit from a UDC sale?
Year to year, we don't see much difference. But this comparison, across three years, does highlight some significant changes. The UDC loan book has grown by 35%, well up on the ANZ.NZ parent growth of just 3.9%. Have ANZ taken the opportunity to guide some of their less desirable loans into the UDC disposal bin? In gross dollar terms, all significant sectors at UDC are higher, except for Finance & Insurance and Manufacturing. Could the fact that ANZ have seen fit to keep more of 'Finance & Insurance' and 'Manufacturing' 'in parent house' mean that this is where the ANZ sees the best growth opportunities for NZ, and itself, going forwards? Returning to the UDC rubbish bin theory, by far the biggest increase in the UDC portfolio was 'personal loans and other lending'. 'Personal loans and other lending' is a very large catch all bucket. I guess a lot of that could relate to small business lending, where proprietors put up their own home as collateral. I wonder if ANZ see this lending as relatively unprofitable and/or difficult to administer? If you look down the 'All ANZ.NZ' column they are now doing $10b less of it.
SNOOPY
What is a 'Stressed Loan'? For the purpose of this discussion, I have a special definition.
Stressed Loan Definition UDC Heartland 1/ Take loan total from categories 7 and 8 a/ Take loans at least 90 days past due. 2/ add 'Default' loans b/ add Loans individually impaired. c/ add Restructured assets. (*) 3/ less Provision(s) for Credit Impairment{s) d/ less Provision(s) for Credit Impairment(s) 4/ equals 'Total Stressed Loans' e/ equals 'Total Stressed Loans''
(*). (Note that from FY2017 ' Restructured Assets' are now not reported on separately by Heartland.)
A 'Stressed Loan' can be thought of as a kind of 'Vulnerable Loan', as previously described (my post 352), but with the Impairment provision taken out. There is no overlap between a 'stressed loan', as defined here, and the amount of money written off each year in bad debts. But yes, 'Stressed Loans' are very much a judgement call by management.
They may
1/ recover,
2/ stay stressed or
3/ have to be impaired and later written off.
As a shareholder in either ANZ (owner of UDC) or Heartland:
1/ I would hope that management would have a robust process that identifies problem loans before they have to be written off.
2/ So as a shareholder, I would hope such loans were seen as 'stressed' before being classified as 'impaired' and certainly before an actual write off was declared.
How does one check that this is what happens in reality? One way could be to look at the 'stressed loans' for both companies on an annual trending basis and see how this compares with the equivalent annual trend in write offs.
Heartland
The column (W) lists the actual dollar amount in bad debts written off over that period, as detailed in AR2018 note 20e.
The key point to note here is that the 'impaired loan expense' / 'write offs' (represented by letter 'W' in each case) only occur:
1/ when the impaired portion of the loan has gone through the whole loan review system'. AND
2/ when a loan repayment has been missed, or a non payment is imminent
Heartland Date 'Stressed' Loans on the books (X) Net Financial Receivables (Impairments deducted) (Y) (X)/(Y) Write Offs (W) Gross Financial Receivables (Z) (W)/(Z) EOFY2013 $48.074m $2,010.376m 2.39% $6.679m+$1.081m $2,060.867m 0.377% EOFY2014 $43.354m $2,607.393m 1.66% $35.258m+$3.260m $2,631.754m 1.46% EOFY2015 $32.824m $2,862.070m 1.15% $1.555m+$1.910m $2,893.724m 0.120% EOFY2016 $32.894m $3,113.957m 1.06% $12.010m+$6.653m $3,140.105m 0.594% EOFY2017 $34.490m $3,545.896m 0.973% $2.140m+$9.531m $3,575.613m 0.327% EOFY2018 $43.278m $3,984.941m 1.06% $4.546m+$14.924m $4,017.436m 0.485%
Note: During FY2016 'Heartland New Zealand Limited' and 'Heartland Bank Limited' combined into a single listed entity.
UDC
I have redefined the 'Total Financial Assets' as listed in note 10d to be 'Net Financial Receivables', because they have already had their 'Provision for Credit Impairment' deducted (netted off).
UDC Date 'Stressed' Loans on the books (X) Net Financial Receivables (Impairments deducted) (Y) (X)/(Y) Impaired Asset Expense (W) Gross Financial Receivables (Z) (W)/(Z) EOFY2014 $95.364m $2,344.131m 4.07% $15.333m $2,375.936m 0.645% EOFY2015 $82.267m $2,429.695m 3.39% $11.503m $2,461.224m 0.467% EOFY2016 $85.475m $2,655.841m 3.22% $9.753m $2,684.750m 0.363% EOFY2017 $116.131m $2,975.781m 3.90% $10.558m $3,005.059m 0.351% EOFY2018 $93.502m $3,283.630m 2.85% $8.875m $3,318.198m 0.268%
Note: The 'impaired asset expense' in the table above is NOT the same as the 'credit provision charge' in the UDC income statement. The former is the cash taken off the books in a calendar year because certain impaired assets have been completely or partially written off. The credit provision charge is an annual adjustment to the 'provision for impairment'.
Discussion
I will preface this discussion by saying that, in previous years, I have made a bit of a hash of things in my tabulated calculations above. Some of this hash was because I have now changed my mind on what ingredients made up the final number. But my major mistake was made looking at UDC There I added "Collective Provision charge to the Statement of Comprehensive Income" (effectively a cash charge as I see it now) to the 'Individual provision bad debts written off' (also a cash charge) without changing the sign of the latter. The latter had a negative sign in front of it - because it was shown in the 'Provision of Credit Impairment' Note as reducing a provision. Absolutely correct, nothing wrong with that presentation. But I wanted to use that figure in a different context of 'cash movement'. 'Individual provision bad debts written off' represents cash out during the year and so does any charge made to the 'Statement of Comprehensive Income' for the year. I should have added these two 'cash out' numbers together, and now I have done just that, I should conclude by saying that, despite these errors, the general thrust of my previous argument has not been affected.
I ended last year's discussion expressing my doubts about the ever decreasing write-off rate at UDC, while the 'stressed loans' did not show such a trend. We are now at the end of FY2018 and the write off rate has dropped again, although this time the stressed loan count is down as well. We mammals like to look for correlations. But I am going to put forward an alternative explanation as to why the 'stressed loans' and 'write offs' should not be correlated. Suppose, as the owner of UDC, you wanted to sell it and were keen for the business to be marketed in the best possible light. Would it not be sensible to really cast your eye more closely that usual over the stressed loan portfolio? That way you could bring about more early interventions to make sure not as many 'stressed' loans became 'distressed'. The counter argument to that is: Why would you wait for a whole of business sale to implement a best practice policy? And isn't it equally important to focus on stopping the unstressed part of the loan book becoming stressed at any time? Wouldn't any decent manager do all this anyway? The UDC sale, in one form or another, was very much on the table during FY2018. So I think it is still too early to say if the lower write off expenses at UDC can be sustained. My critical eye remains on UDC in this regard.
In the case of Heartland, the 'stressed loan' percentage went down with a thump as Heartland extracted itself from its legacy property problems over FY2013 and FY2014.. Stressed loans have reached a plateau of about 1%. The actual write offs at Heartland were very high in FY2014, and that year should be seen as an outlier. There is a pattern of up and down years (highs and lows correcting each other over time?) averaging some 0.45%. This contrasts with the monotonic decline at UDC
The UDC stressed loans look floats around at 4% of the total. This is much higher than Heartland. But this could be due to the nature of the business rather than management incompetence. UDC loans are perhaps more heavily weighted towards 'working equipment' which may not have much 'fire sale' value during a business downturn. Yet I would have expected more diligence from Heartland, simply because they have more staff. The Heartland pattern of reducing 'write offs' coupled with reducing 'stressed loans' does make a more plausible narrative than what is happening at UDC.
SNOOPY
Big day for Aussie banks today with the Royal Commission Report coming out
Some say the shorters of banking stocks could get hurt
And yet , from the AFR
Opinion
Shorting the big four banks: widow-maker no more?
https://www.afr.com/news/economy/sho...0190203-h1asf7 (only the first paragraph for free)