This sector is fascinating. Every time I read a note, speak to an analyst or look at a model there seems to be a different view or emphasis. The variances in estimates and price targets are huge. There must be an opportunity here.
Because banks are so opaque and difficult to analyse my current focus has merely been to try and establish where the key points of contention are. Once this is done it should be a lot easier to establish whether or not there is genuine value here. As far as I can make out there are basically three areas of disagreement: pre-provision profits; total impairments; and the starting capital base.
1) Pre-provision profits: In a bull note on RBS published yesterday, a well known UK broker appears to be constructing a bull case for RBS on the basis of higher pre-provision profits than what the bears are assuming. For the 2008-11 period they have a total of £48bn of cumulative impairment, versus the uber bears who are assuming £49.6bn, slightly higher but not much difference. They key difference is the level of cumulative pre-provision profits - £35bn versus only £26bn for the uber bears. This constitues an extra £10bn of capital. Interestingly, this bull case does not rely at all on the other bull argument in the market - the assumption of lower impairments on the non-APS book versus the APS book (see 2 below). As stated explicitly in the note today: “With so few details on the composition of the assets that will be covered from APS, we do not assume any reduction in the impairment charge, once RBS has utilised the £19.5bn first loss”.
2) Impairments on the non-APS book: The other source of bullishness for the sector is the argument that the non-APS book will experience much lower provisions. Most analysts (bulls and bears) don’t assume anything for this, instead using a normalised number, although some of the bulls recognise that this could become a source of upside once more information is available. It is interesting that some analysts’ do believe that they already have sufficient information to make the assumption of lower non-APS impairments versus APS impairments. This is clearly a really important issue, and potentially one that the company's themselves may be able to provide some colour on.
3) Deductions to capital: The uber bears on the sector depress their starting capital by subtracting up front the non-amortised portion of the APS fee and the EV deduction for the life business. Others don't do this. Having considered this issue, I have concluded that the bearish view is illogical, at least as far as the APS fee is concerned. The bear arguement is that the APS fee is an intangible asset and should therefore be deducted from equity as you would goodwill. The problem with this analysis is when inferences are then made as to what the ROE should be, on the basis of the written down capital base. The APS fee represents payment for a long term insurance contract on the banks’ assets. If BP were to purchase a 7 year fire insurance contract on their head office, I would not deduct the entire cost immediately from the shareholders equity. And even if I did, I certainly wouldn’t make inferences about whether or not the resulting ROE was sustainable (the ROE would be higher because equity would be lower, but this would clearly be an accounting illusion). The APS fee should be considered as a cost of running the business and should therefore be spread over the life of the contract, ideally reflecting the timing of the losses.
That said, even if I do adjust the bear's numbers by adding back the APS fee to capital, the resulting NAVs are still some way lower than where the bulls are (especially when you get out to 2011 and 2012 when most of the fee has already been amortised). This suggests to me that a bull case on these stocks probably rests more on points 1) and 2) rather than 3). This means taking a view on how bad you think the UK economy is going to be over the next 2-3 years relative to what is already being assumed and, crucially, the extent to which LBG and RBS have managed to dump all their toxic assets into the APS. I suspect the investment case would need to rest more on the latter point rather than the former.
Friday, June 26, 2009
Wednesday, June 24, 2009
UK Banks: no longer a mug's game?
I am convinced that UK domestic banks are least understood sector in the market, despite all the media and analyst coverage. The basic reason for this is the huge divergence in forecasts for impairments, NAVs and normalised earnings. As an example, the list below shows the forecasted NAV/share for Lloyds Banking Group (LBG) from various sources.
HSBC: LBG NAV (2011) = 112p
RBS: LBG NAV (2011) = 92p
Redburn: LBG NAV (2011) = 99p
Panmure Gordon: LBG NAV (2011) = 74p
JP Morgan: LBG NAV (2011) = 20-25p
Now this is a most unusual situation. Normally analysts huddle together, unwilling to take a stand too far in one direction. Situations such as this, where this is a huge variance of forecasts make for potentially very attractive investments, assuming you can get your analysis correct. I don’t have the answers yet, but I am finding myself becoming more positive on both the UK domestic banks. There are two main reasons for this:
1) A misunderstanding of the asset protection scheme (APS). The way the APS works is that both LBG and RBS have to take the “first loss” on their asset impairments (£25bn pre tax in the case of LBG), after which the Government takes 90% of any further losses. However not all of the banks’ assets are in the APS. The idea was that only the bad stuff would be put there, in effect creating a pseudo good bank/bad bank structure. A really important question from a valuation perspective is what will happen to the non-APS assets – roughly £460bn in the case of LBG (£260bn are covered by the APS). Many of the bears appear to be assuming continued high level of impairments on the non-APS assets into 2010 and 2011 once the first loss is hit (it is consensus that LBG will hit the first loss in early 2010). However, a contrarian on the sector recently made an excellent point to me: the losses on the non-APS assets are actually likely to be substantially lower than on the APS assets, much lower indeed than most people appear to be assuming. Many of the bears appear to be assuming similar or marginally lower level of losses on the non-APS assets. He reckons the non-APS assets are likely to experience around 25% of the losses experienced by the APS assets. If you run these numbers then the trough valuation looks much more appealing.
2) Economic versus accounting adjustments - NAV / RoE circularities. The other issue with banks, one that becomes much more prevalent during recessions and bear markets, is capital and reserves. During a boom no one cares about book value, but once asset impairments enter the picture suddenly all anyone can think about is whether the bank has enough capital. While this is important from a regulatory perspective, there comes a point where the economic reality becomes obfuscated by accounting gimmicks. For example, some of the bearish analysts are now deducting from capital the APS fee, and doing it up front. Since this is £15.6bn it seriously reduces the capital base. Then they say, “look how high the ROE is! that cannot be sustainable, the bank must need more capital”. The ROE is of course much higher because the capital base has been reduced by 30-40%, but does this actually matter? Is it economic? Imagine if a normal company were to announce that they were paying all their staff 7 years salary up front. Would it make sense to deduct this from their capital? Of course not, it should be amortised over the whole 7 years. Once you realise this you can see that conclusions based on implied ROEs have to be treated with much scepticism.
This idea is still very much work in progress. Clearly there are still huge problems with the UK economy, so impairments on the non APS book could in fact be higher than the bulls are assuming. It is also still not entirely clear how risky the assets outside the APS book actually are. Nevertheless, preliminary work suggest an NAV per share for LBG of 100-120p could be achievable by 2011. And when you consider that this is likely to mark the bottom of the impairments cycle and the bank will have a quasi monopolistic position with the domestic mortgage market, it is not hard to imagine a 1.5-2X multiple of this NAV, implying 175-250p potential value.
HSBC: LBG NAV (2011) = 112p
RBS: LBG NAV (2011) = 92p
Redburn: LBG NAV (2011) = 99p
Panmure Gordon: LBG NAV (2011) = 74p
JP Morgan: LBG NAV (2011) = 20-25p
Now this is a most unusual situation. Normally analysts huddle together, unwilling to take a stand too far in one direction. Situations such as this, where this is a huge variance of forecasts make for potentially very attractive investments, assuming you can get your analysis correct. I don’t have the answers yet, but I am finding myself becoming more positive on both the UK domestic banks. There are two main reasons for this:
1) A misunderstanding of the asset protection scheme (APS). The way the APS works is that both LBG and RBS have to take the “first loss” on their asset impairments (£25bn pre tax in the case of LBG), after which the Government takes 90% of any further losses. However not all of the banks’ assets are in the APS. The idea was that only the bad stuff would be put there, in effect creating a pseudo good bank/bad bank structure. A really important question from a valuation perspective is what will happen to the non-APS assets – roughly £460bn in the case of LBG (£260bn are covered by the APS). Many of the bears appear to be assuming continued high level of impairments on the non-APS assets into 2010 and 2011 once the first loss is hit (it is consensus that LBG will hit the first loss in early 2010). However, a contrarian on the sector recently made an excellent point to me: the losses on the non-APS assets are actually likely to be substantially lower than on the APS assets, much lower indeed than most people appear to be assuming. Many of the bears appear to be assuming similar or marginally lower level of losses on the non-APS assets. He reckons the non-APS assets are likely to experience around 25% of the losses experienced by the APS assets. If you run these numbers then the trough valuation looks much more appealing.
2) Economic versus accounting adjustments - NAV / RoE circularities. The other issue with banks, one that becomes much more prevalent during recessions and bear markets, is capital and reserves. During a boom no one cares about book value, but once asset impairments enter the picture suddenly all anyone can think about is whether the bank has enough capital. While this is important from a regulatory perspective, there comes a point where the economic reality becomes obfuscated by accounting gimmicks. For example, some of the bearish analysts are now deducting from capital the APS fee, and doing it up front. Since this is £15.6bn it seriously reduces the capital base. Then they say, “look how high the ROE is! that cannot be sustainable, the bank must need more capital”. The ROE is of course much higher because the capital base has been reduced by 30-40%, but does this actually matter? Is it economic? Imagine if a normal company were to announce that they were paying all their staff 7 years salary up front. Would it make sense to deduct this from their capital? Of course not, it should be amortised over the whole 7 years. Once you realise this you can see that conclusions based on implied ROEs have to be treated with much scepticism.
This idea is still very much work in progress. Clearly there are still huge problems with the UK economy, so impairments on the non APS book could in fact be higher than the bulls are assuming. It is also still not entirely clear how risky the assets outside the APS book actually are. Nevertheless, preliminary work suggest an NAV per share for LBG of 100-120p could be achievable by 2011. And when you consider that this is likely to mark the bottom of the impairments cycle and the bank will have a quasi monopolistic position with the domestic mortgage market, it is not hard to imagine a 1.5-2X multiple of this NAV, implying 175-250p potential value.
Subscribe to:
Posts (Atom)