Saturday, October 1, 2011

Time for Discover

Financial stocks in the US have been very weak.

The consensus is that the situation is bad on credit. I suspect the situation is more likely to be bad on revenue - but either way the consensus is bad for financials.

I am going to give you an anti-consensus set of charts. This is credit data on Discover Financial Services (NYSE:DFS) from their master-trust for their credit card products. The information is from indispensable Portales Partners. They do not like people redistributing their stuff - but I hope they are happy with an advert: if you run a serious amount of money in North American Financials you should subscribe to them. They are likely to make you think.*

Here is the last ten years of DFS delinquency data:




Note that delinquency is near a ten year low.

And here is the charge-off data:




The spike and the collapse in the numbers in the middle of the sequence is the change in personal bankruptcy laws which gave people an incentive to bring forward bankruptcy filings. This meant a spike and subsequent drop.

Net of the spike which was caused by a policy change the charge-offs are about the lowest in a decade.

Delinquency leads charge-offs so given the delinquency is low you would expect the charge-off to drop.

Some of the low charge-offs are caused by high recovery of past written-off debts (recoveries count as negative charge-offs).


Recoveries have come back to a high level.

Observation

DFS are reporting among their best credit numbers in a decade - and they are reporting it in a very sour economy. If the delinquencies really are a leading indicator these numbers will be the best in a decade shortly.

So what is happening?

I can see three broad possibilities:

(a) The numbers are not real - DFS is faking it.

(b). The numbers are real but they are DFS specific and they relate to changes in DFS policy such as a dramatic tightening of credit standards, or

(c) The numbers are real and middle American unsecured credit has improved dramatically despite the recession because consumers have retrenched and really are paying back their loans.

I will leave it to my commentators - you are a smart and well connected lot - to work out which.

But if it is (a) then DFS is a strong sell, and if it is (b) DFS is a strong buy. If it is (c) then there are a wide range of financial stocks you should buy.




John

*Seriously - this advert is warranted. Portales really are one of the better specialist firms out there.

Post Script

I know I am cheating with this list of choices - there are other possibilities like

(d) It is all going to get worse ... so don't buy financials when this data is at its peak (in other words do not buy).

(e) DFS earnings are artificially inflated with recoveries and hence earnings will fall when recoveries normalize (in other words do not buy).

(f) The revenue line for all American financials is toast as per the Japanese experience. So far one of the biggest problems for American financials has been falling revenue but that is not widely commented on outside specialist bank analysts. (In other words sell.)

I left it vague because I wanted to encourage comment. I learn a lot from the comment - but I got many emails that thought I was being unfair and simplistic. So I needed to clarify.

Let me tell a story though.

When I first went to visit BofA in Charlotte (and this dates me) I wanted to ask them about credit. They told me they had 36 billion in revenue and 18 billion in costs and 2 billion in credit costs and that I should watch where the revenue and cost lines were going because they drove it.

That turned out to be wrong of course.

But it may not be wrong in the future.

Post Script 2

I think I should say this is not Discover specific. Consumer credit looks about as good (ie safe) an asset as it has ever been. (It can get worse..) It can also be spurious (extend and pretend for instance).

The best comment yet received by email suggests that I should measure the delinquency against availability of cheap rollover credit cards. Why default - or so the argument goes - when someone will give you a balance transfer and extended credit at 2 percent?

25 comments:

Anonymous said...

John,

A few guesses:

In Credit Card Asset Backed Securities, you can monitor the performance of accounts in the Trust and remove an account when it becomes deliquent. There weren't enough good accounts in '07-'09 to keep chargeoffs from spiking. This is the main difference between ABS and MBS.

Also, of all the Credit Card issuers, Discover has grown the least BY FAR. They have basically chose to not grow their portfolio at all in the last several years.

C or BAC Trust data might be a better indicator.

Regards,

AK

deaner said...

How does DFS define delinquency - is it just carrying a balance (I would doubt that - that's the CC issuer's business model); carrying a balance without making the minimum payment (more plausible), or; carrying a balance without a series of minimum payments (a little aggressive for my tastes, but plausible). It doesn't really matter - if the definition hasn't changed it's just the trend we are looking at, but I am curious. Also, the dip in charge-offs after the change in BK laws looks deeper and longer than just a recovery after filings were advanced into the pre-change peak; I don't think they are 'lowest in a decade' - but they are heading in that direction.

FWIW, I think scenario 3 is most likely - but other than implicitly (by not hoarding gold, canned food, and ammo) I'm not putting any money on it.

Cheers,

Dean

Jes said...

Or d) it's real and about to turn for the worse.

Anonymous said...

You are missing the real issue (which surprises me since I know you know how to analyze financials). The real issue is the provisioning and reserve release, the latter of which has been driving earnings at DFS. Core EPS are about 40c per quarter lower (e.g. if provisions matched the losses). With delinquencies at a 25 year low (check the Q2 call transcript) and provision to revenue at 6% (another all time low), provisions are going to HAVE TO increase to match chargeoffs (they admitted this on the call). EPS are likely to be down on the order of 30% next year, and if the loss rate begins to increase (quite possible), DFS will earn even less. It's a sell. The wild card of course are reserves which remain substantial at 440bp / loans (around 1 year of "normal" losses). Maybe they bleed that down a little over 2013-14, but near term the reserve release story is OVER.

Anonymous said...

Why are stocks are buy if you have scenario C?

Scenario C would mean weak revenues, operating leveraging, and may be correlated with tighter NIMs among a wide variety of financial institutions... right?

Anonymous said...

In fact, would any one of the scenarios lead to a strong buy recommendation?

Even in scenario B, the problem will be revenues, not credit.

The charts you show are all consistent with a deleveraging "Japanese" scenario. As you have pointed out before, Japanese credit quality is very good. Borrowers and banks are both very conservative. The problem is and will remain revenues.

H.Z. said...

Your list of choices are false choices. The numbers are real, but that does not mean DFS or financials are strong buy in general. DFS is a weak franchise that is trading at a significant premium to book. Its performance comes from the fact that fear has pushed most financials to tighten up on unsecured credit and low interest rate masks its weakness in deposit collection. With such good credit performance either lending rate will have to come down or the card portfolio will dwindle away. No good credit risk will pay the high card rates more than temporarily in this low rate environment. Or the corollary is also true: those who pay high card rates in such a low rate environment are just ticking time bombs evidenced by the fact that no other lenders are competing for their borrowing.
However I agree with your general point that the credit card companies are trading a level that does not portend another recession at all (or at least not significant uptick in unemplyment -- so recession or not unemployment probably has peaked). So the other financials are weak for their own reasons (lending margin and legacy mortgage issues).

Anonymous said...

I see you have decided to leave the problem on the board for the class to work out. As a person with a net worth consisting of tennis shoes, my anecdotal experience is that a) seems more likely, as a lot of people I know are having credit problems and people are using cards for basic necessities, which means a lack of resources. I also don't know anyone (at all!) that uses Discover. By the way, every post you feel fit to print is valuable. Thank you. Going to check out that website.

狂猪 said...

Check this out
http://www.federalreserve.gov/econresdata/releases/statisticsdata.htm

In particular look at the Bank Assets and Liabilities section. There is a charting function on the data download page (look for DDP). It is a very nicely done.

I agree financials should continue to improve.

Businesses are sitting on a big pile of cash. The weak ones have already went under. Cost, and in particular staffing, have already been cut to the bone. Therefore, unemployment is not likely to rise either.

Anonymous said...

Perhaps a stupid question, but why are charge offs higher than the delinquency rate? I'd think you charge off only part of the loans that have become delinquent?

Anonymous said...

Delinquencies are low thanks to negative home equity...

Anonymous said...

I think you will see similar trends in delinquency improvement with other credit card companies (AmEx, BAC, etc.). If we're not going to give these entities credit for releasing reserves when they're over-reserved, then logically it follows that we should also not penalize them for taking provisions when they are under-reserved.

Anonymous said...

it looks like the last time delinquencies were this low was august 2007, not an auspicious time to be buying financials

Anonymous said...

John,

I've also sent you an email
--

There are some very simple potential explanations for the trends you posted about Discover. Here is my take:

- Delinquency and charge-offs are inversely correlated with availability of short-term credit. In other words, there is no reason for me (as a consumer) to be delinquent in paying 2% minimum payment every month when I can rollover my existing debt with Discover to another card company and pay 0% or 1% annually for the next two years. Bernanke has kept rates low for the past few years and has promised ultra-low rates for the next two years. Other card companies are flooding consumers with balance transfer checks to roll over debt (without increasing their overall credit availability). Most Discover card holders have Visa/Mastercard from other credit card companies and hence this process is easier.

- Note that this is just the old game of rolling over debt from one firm to another. Consumers have been trained to do so for a very long time in the US. The macro environment has forced card companies even more to play this game. At 0% funding costs, if they can get 2% for two years on tens of billions of dollars, that is still a decent chunk of money for their top line in short-term.

- In fact, consumers don't need to even rollover my debt. They will just write themselves a check using the Visa/Mastercard credit offers from other companies such as Capital One and use that to pay the minimum amount to Discover.

- Credit card companies have two big risks they can't control. 1 ) Unemployment and any drastic increases in it. 2) Spiraling healthcare costs. The second leads to bleeding and blood letting in their portfolios. The first risk just blows a big bomb when it happens with little or no advance notice.

- Discover is not indicative of mainstream US consumers anyway. They cater to a fringe sub-sector of the populace. Extrapolation to US consumers across the board will be faulty at best.

Anonymous said...

Once again, the guy with tennis shoes as his net worth. I think leaving it open allows us to think it through. you normally give your answer then take us through your thoughts. If anything, reading your thought process led me to a). To paraphrase a saying, "If you look only for an honest man, you'll find yourself with no companions." Banks are lending to people with stellar credit. Such people are rare. As credit tightens, such people become extinct. If that is your customer base, you too, die along with your food source. Not immediately, though.

Anonymous said...

Many consumers are now using their credit card for everyday purchases. They protect their credit card (ie the last bill they do not pay) because that is their lifeline to buy groceries, gas, etc.

There are also people that have strategically defaulted on their home. While the foreclosure process is slow, they live rent/mortgage free. No mortgage payment means they can pay their credit card bills and/or save money.

Andrew Simms said...

I have recently been parsing through Fed data, specifically focusing on credit trends in relation to the banking sector. It looks as if C&I as well as consumer loan growth is beginning to accelerate while at the same time net intrest margins are compressing. This sets up an interesting dilema where the only net positive for banks is if loan volume superseds the shortfall from tighter margins. The fact that discover financial has lower delinquency rates and charge offs is pretty much irrelevant when it comes to the health of the financial sector. You should be very bullish on banks right now if you believe the US will avoid a nasty double dip recession and if you believe that loan growth is accelerating which may be the case. Some US baks are trading at significant discounts to tangible book value implying huge write downs and negligent future eps. For myself, I am somewhat of a coward hiding out in Canadian banks as I am unsure where markets are headed in the near term....

Baltazar said...

Let's preface this by agreeing that it's untimely to consider financials, and there are many more promising financials to consider, after we feel the effects of the European liquidation.

One curious thing about DFS is that it doesn't do a good job scrutinizing changes in customers' credit scores as they naturally deteriorate through the Depression.
Other credit card companies would lower available credit balances or even close the accounts when they periodically re-scan for FICO changes.

This is also a counter-argument to the durability of the rollover thesis. Aggregate credit score will continue to deteriorate, almost inversely to future acceptance of rollover candidates.

It's all just a part of the same self-reinforcing deleveraging cycle. We do a terrific job extending and pretending, so it's easy to be in denial.

Anonymous said...

There are also people that have strategically defaulted on their home. While the foreclosure process is slow, they live rent/mortgage free. No mortgage payment means they can pay their credit card bills and/or save money.

I keep reading this explanation, but I have never seen anybody put solid evidence (ie, data) forward to back it up.

Who knows--maybe it's true (WFC's classification of what constitutes 'delinquent' is certainly amusing). But I would really like to see the actual numbers behind this.

Nemo Incognito said...

The senior loan officer survey data looks good too... not denying that there are plenty of reasons one *should* buy banks here but frankly deflation is death for banks and, well, stagnation is what this looks like until at least the middle of next year. Long term value perhaps but this market can't see past Greece's next coupon payment.

The more yogi bear trades here seem to be in securitizations and the likes of NLY versus BAC. Anything reasonably short dated and mezzanine will pay and has been heavily sold off. 15% odd yields might not be a Shinsei type return but the turn will come in this stuff before the banks.

Doc at the Radar Station said...

I believe it's a *combination* of:

(c) The numbers are real and middle American unsecured credit has improved dramatically despite the recession because consumers have retrenched and really are paying back their loans.

(d) It is all going to get worse ... so don't buy financials when this data is at its peak (in other words do not buy).

Justification for c).. People are delevering, no doubt about it. We've definitely got a case of credit revulsion. I'm just a median income guy with a bachelor's degree working in a manufacturing environment and talk with the full spectrum of income earners every day and you hear a LOT of talk about "getting out of debt".

Justification for d).. It looks like a pattern that follows the business cycle and it's fixing to roll over. Growth rates have slowed, and all it's going to take is another crisis and unemployment will probably lurch from 9ish up to 11ish (or worse) and the default cycle will start again.

Twist said...

The revenue-maximizing borrower pool is not one with great credit because card accounts lose money on borrowers who pay off their balance each month. Of course a pool with terrible credit is not great either because of high chargeoff rates, but due to the changes in bankruptcy law you mentioned, a pool with worse credit quality can be more valuable.

So my read of the data is negative for DFS - given that the unemployment rate has been flat since June 09, the falling delinquency and chargeoff rates indicate that the borrowers that got into trouble during the initial spike in unemployment have left the pool. The remaining borrowers are likely to have solid jobs because they've been able to survive the negative shock. If you have a chart of mean FICO over time for the non-delinquent fraction of the pool, that would be interesting.

Anonymous said...

Not a financial gury but it would seem to me that financial companies retrenched in 2009 and 2010 and concentrated their business on the best credit (that's why they stopped sending out credit card solicitations). Which leads to a low delinquency rate. But which also leads to stagnated revenues.

Credit card companies made their hay those with bad credit through fees, late charges, etc. I'm not sure there is a business model where they lend solely to people who pay off their balances every month.

Anonymous said...

Take a look at Tangible Book Values and how they have been increasing year over year through the second quarter. If the banks were in worse shape I think they would (should?) write down more assets and show a smaller increase?

Below were given by the companies in their filings.

AIG 8.7% YoY 2Q
BAC 4.2% YoY 2Q
C 16.4% YoY 2Q
GS 7.8% YoY 2Q
MS 8.5% YoY 2Q

Below were calculated by me as the companies did not provide the numbers

JPM 7.0% YoY 2Q
WFC 14.5% YoY 2Q

In regards to foreclosure/rent free comment I made previously. I live in Phoenix, AZ and definitely have heard it and seen it. Not sure if there is any data out there about it. I remember reading a few articles about it but can't find them linked anywhere online.

Anonymous said...

John (or anyone):

Any idea of how much Portales Partners charges for access to their research?

Thanks.

General disclaimer

The content contained in this blog represents the opinions of Mr. Hempton. Mr. Hempton may hold either long or short positions in securities of various companies discussed in the blog based upon Mr. Hempton's recommendations. The commentary in this blog in no way constitutes a solicitation of business or investment advice. In fact, it should not be relied upon in making investment decisions, ever. It is intended solely for the entertainment of the reader, and the author.  In particular this blog is not directed for investment purposes at US Persons.