DISCOVER FINANCIAL SVCS INC DFS
January 04, 2012 - 5:39pm EST by
compass868
2012 2013
Price: 24.31 EPS $4.06 $4.30
Shares Out. (in M): 533 P/E 6.0x 5.7x
Market Cap (in $M): 13,000 P/FCF 0.0x 0.0x
Net Debt (in $M): 0 EBIT 0 0
TEV ($): 0 TEV/EBIT 0.0x 0.0x

Sign up for free guest access to view investment idea with a 45 days delay.

  • credit card
  • Consumer Finance

Description

DFS is a good long here.  Fair value is low 30’s, and stock will get there based on material upcoming earnings revisions.  This is a squeaky clean financial, with exceptionally strong capital, and earnings that are very underestimated by the street.

Business

DFS is primarily a credit card issuer with $46b in card receivables.  They also have a small and growing private student loan business ($7.3b in receivables), and a debt consolidation business ($2.6b in receivables).  The student loan business is mostly from an acquisition in q1 2011 from Citi, where DFS marked the book down meaningfully and is exposed to limited credit risk as a result.  DFS also owns their own payment network (only one of four networks, the others being MA, V, and AXP).    Finally, DFS has a 3rd party payments business which mostly processes pin debit transactions.  Payments is small (only 4% of revenue) but potentially strategically valuable.   

The company somewhat avoided the financial crisis by not growing its card receivables during the boom credit years of ‘06-08, and therefore avoided a lot of troubled credits underwritten by large issuers.  As a result they only reported modest operating losses in two quarters during 2009.  DFS has a sticky customer base, growing wallet share, and growing merchant acceptance.  Merchant acceptance has grown from 76% in 2006 to 97% in 2010, which bodes well for DFS share of spend going forward. 

Long case

Street is underestimating earnings due to three main factors – 

1) Funding costs.  DFS currently funds all in at 2.60%, composed of $25b in CDs at 3.30%, $11.1b in money market accounts at 1.26%, $15.5b in ABS at 2.05%, and $2.6b long term debt at 6.43%.  I believe they will optimize their funding mix over the short and intermediate term which will provide a tailwind to EPS not currently captured by the street.  Most obviously, their CD book matures $2.6b in 2011, $10.8b in 2012, $5.9b in 2013, and $6.2b thereafter.  Their current weighted maturity CD offer rate is ~1.33% (you can find the rates on their website) and I am assuming a conservative average maturity of nearly 3 years although I believe they target closer to 2 years.  So their marginal cost of CD funding is1.33%, or 60% lower than their existing CD funding of 3.30% I mentioned earlier.  As such they will refi more than half the CD book by 2012 at less than half the current cost.  Secondly, their savings rates are about 0.30% higher than competitors and they will bring this down as well (move from 1.26% to 1.00%).  In addition, they are growing their savings funding rapidly which will replace their reliance on more expensive ABS funding.  All in by 2012, this mix shift should result in a funding benefit of ~60 bps or $0.40c to EPS which I don’t think is captured by the street. 

2) Credit costs. 

Credit costs have been rapidly declining for all card companies and DFS in particular peaked at 9% in q12010 and was only 3.24% in q42011.  Credit costs are coming down for several reasons.  First, although aggregate unemployment is high, the net inflow to unemployment is stable (and has been for some time), and it’s really the event of job loss that drives someone to charge off.  Secondly, industry underwriting has been overly conservative the past several years.  Said another way the loan book is seasoned and most people who will charge off have already charged off.

It’s commonly understood that in a normal cycle (i.e., not a recession) credit card charge offs will be ~ 5%.  The street is modeling ~3.70% for DFS in 2012 and 4% in 2013, i.e. trying to adhere to the company’s/industry’s comments on the 5% number.  I think however, that we will enter a multi year period where we could meaningfully overshoot credit costs, and the market has not fully realized that and street EPS estimates will have to be revised higher.

The charge off data is trackable and is reported monthly in the issuer trusts.  30-59 day trust delinquencies have been falling dramatically (driving the initial credit improvement we have seen), but work into charge offs on a lagged basis.  The “roll rate” from the 30-59 day delinquent bucket into the charge off bucket has been pretty consistent over time (in general 75% of those who are 30-59 days late move into the 60-90 day DQ bucket, 85% of 60-89 days DQ roll into 90-120 days DQ, etc….until balances are automatically charged off at 180 days past due.)  On average the conversion rate from 30-59 past due to charge off is ~37% net of recoveries. 

DFS’ most recent trust data was reported for November at a 3.04% net charge off rate.  That charge off bucket flowed from the cohort of 30-59 DQs in June 2011 which was 72 bps.  Since June 30-59 day DQs have continued to decline to 69 bps as of November.  So for at least the next 6 months charge offs should be in the 2.90% range.  In addition, the last two reported months saw a spike in the roll rate from the 150-180 day DQ bucket to the NCO bucket.  In October and November the roll rate from 150-189 to NCO was 100% and 98% respectively, versus the trailing 12 month average of 92%.  That 92% has been pretty constant over time and I believe it will mean revert.  I think the last two months were driven by exogenous factors (random spike in bankruptcies, estates entering probate etc).  With dollar delinquencies as low as they are even a small move in the roll rate can affect NCOs.  In any event with a 92% roll rate to NCO, NCO’s should be ~ 2.70% for at least the next six months.

Delinquencies are still falling but at a lower rate than they have been, but it’s worth noting that they fell consistently over the past six months which is usually a seasonally up period for DQs.  In the next six months, the normal seasonal pattern is lower which would argue for NCOs  lower than the 2.70% in 2h12. 

As far as losses into 2013, the loan book is not growing rapidly (a few % per year), and normal peak charge offs on new originations are usually reached 18-24 months post origination.  Therefore, even as DFS grows its book modestly in 2012, it should be 2014 before that book starts to season and credit costs go up.

The other way NCOs could go back up would be a massive spike in new unemployment claims.  This is not happening (in fact the opposite) at present, but double dip risk is out there of course.  That said, a sell side firm published a regression of the relationship between the change in new claims versus the change in delinquencies.  Usually delinquencies move up at roughly half the rate of new claims, but it can be volatile in the tails so let's say it's worse and it’s a 1:1 relationship.  So if claims go up 30% from 381k/week currently to 500k/week (implying a huge negative move in the economy and a claim number not seen since the beginning of 2010), DFS NCO’s could go from 2.70% to 3.50% (up 30%).  Barring this substantial retrenchment, I believe we will stay in a prolonged period of abnormally low charge offs.  That said, even if that does happen, the street is at 4% NCOs in 2013! Every 50 bps of NCOs is 35c to EPS.

So by looking at the current 30-59 day delinquencies, not assuming much improvement from here, but using roll rates which have proven pretty stable over time, it seems DFS should report ~2.70% NCO in 2012 and perhaps 3.00-3.40% in 2013 versus consensus of ~3.60% and 4.00% in both years. 

3) Revenue suppression.  DFS records contra revenue to account for interest and fee revenue that it will not collect based on its current delinquencies.  On an annualized basis in Q3 this cost them $605m in revenue or $0.70 in EPS.  As delinquencies migrate down this revenue “headwind” will go away as well.  It won’t go to zero (there will always be DQs and charge offs), but if charge offs go to 2.7% as I expect from Q3’s 3.85%, this would be a $180m revenue benefit or $0.21c to EPS in 2012.  Note, I am not baking this in for conservatism, because I think they will have natural yield compression in their book from increased promotional balances (lower APRs), the continued reduction in high rate balances (from delinquent customers that pay the penalty rate but can no longer be repriced due to the CARD ACT), and an increase in customers who pay their balance each month (due to the improving economy).  However, should these other yield compressions prove less onerous, revenue suppression could provide upside. 

Valuation.  The current consensus estimates are $3.35 in 2012, and $3.37 in 2013.  With lower funding costs, sharply lower losses, and no reserve releases (in fact some build in 2013), DFS should report ~$4.30 in both years.  With these sort of positive earnings revisions I think the stock can conservatively move to the low $30s (8x peak earnings of $4.30).  There will be some recognition that DFS is “overearning” because the market will think charge offs are unsustainably low.  However, given the time it takes new originations to season and that we could be in the < 4% loss range for the next couple years, I think the stock can still work higher.  Even at a 4% NCO ratio, EPS would be $3.50+ which would support a low $30’s price (and in the interim they will be earning more than that).                      

Conservative balance sheet with significant excess capital.

DFS’ targeted TCE/TA ratio is 8%, and they are currently at 11.4% which implies $2.3b of excess capital.  DFS generates ~$2.0b of capital per year, so by the end of 2012 they should have ~$4.0b in excess capital (less than the simple addition due to factoring in asset growth) or 31% of the current market cap which will be available for buyback or acquisition.

Recently the company announced a $1b buyback (8% of shares) which speaks to their prudent capital management.  They bought back 3% of the company in the past two quarters. 

DFS has significant liquidity with $10b in cash and treasuries (15% of assets) yielding only ~70bps.  They are in the process of investing this liquidity which can be modestly additive to EPS. 

They are a disciplined acquirer with low risk of a large, lateral acquisition.  There are limited card or student lending assets in the marketplace and they do not want to get into other areas of banking (mortgage, auto). 

Modest growth.  Card balances are bottoming after 10 quarters of decline from consumer deleveraging.  All the major issuers have shown 2h11 growth.  This is one area where card stocks are differentiating themselves slightly from banks which have pockets of growth (commercial and industrial, card, auto) but have significant run off books (home equity, construction, commercial real estate) that are making their aggregate loan growth zero or negative.  The growth here is modest (low single digits) but it’s still relatively better which should help continue to make these stocks attractive to investors.

Limited regulatory overhang.  The card companies are one of the areas in financials that don’t have Dodd Frank bill overhangs.  The CARD Act was enacted in 2010 and has been mostly absorbed.  Durbin bill only impacts debit, not credit.  Credit interchange is unlikely to be legislated given how hard it was for Durbin’s debit bill to pass.  The CFPB is mostly focused on disclosure for card companies and will likely attack mortgage before it would seek to do anything else in card.

Company doesn’t need to exist.  Credit cards are still an attractive business if done properly (highest ROA in financials).  As such DFS could be attractive target for a buyer.  DFS is an odd duckling in that it’s the only monoline card company aside from AXP (COF is another but it’s almost half a bank now).  Also, DFS’s payment network has strategic value because it could be used to internalize one of the major issuers spend volumes away from MA/V (JPM or BAC could own their own network by buying DFS and avoid paying fees to MA/V).  Although I think a major issuer buying DFS for its network is unlikely, I do think at some point when we get more capital clarity over the next 6-12 months, DFS has the potential to be bought as a way for a buyer to expand in the card space.

Risks -

Crowded.  Card stocks have been good performers YTD due to the cleanliness of the story relative to other financials, lack of regulatory overhangs, good credit performance, and positive earnings revisions.    

Timing of revisions.  I’m not sure when the street will really revise to the levels I think they need to.  I can’t make sense of their existing assumptions.  Further revisions should come as we continue get master trust data each month.  Eventually, I think you will get sharp revisions and continued beats that filter into numbers and the stock should follow.

Catalyst

Material earnings beats and revisions.  Continued capital deployment.
    sort by    

    Description

    DFS is a good long here.  Fair value is low 30’s, and stock will get there based on material upcoming earnings revisions.  This is a squeaky clean financial, with exceptionally strong capital, and earnings that are very underestimated by the street.

    Business

    DFS is primarily a credit card issuer with $46b in card receivables.  They also have a small and growing private student loan business ($7.3b in receivables), and a debt consolidation business ($2.6b in receivables).  The student loan business is mostly from an acquisition in q1 2011 from Citi, where DFS marked the book down meaningfully and is exposed to limited credit risk as a result.  DFS also owns their own payment network (only one of four networks, the others being MA, V, and AXP).    Finally, DFS has a 3rd party payments business which mostly processes pin debit transactions.  Payments is small (only 4% of revenue) but potentially strategically valuable.   

    The company somewhat avoided the financial crisis by not growing its card receivables during the boom credit years of ‘06-08, and therefore avoided a lot of troubled credits underwritten by large issuers.  As a result they only reported modest operating losses in two quarters during 2009.  DFS has a sticky customer base, growing wallet share, and growing merchant acceptance.  Merchant acceptance has grown from 76% in 2006 to 97% in 2010, which bodes well for DFS share of spend going forward. 

    Long case

    Street is underestimating earnings due to three main factors – 

    1) Funding costs.  DFS currently funds all in at 2.60%, composed of $25b in CDs at 3.30%, $11.1b in money market accounts at 1.26%, $15.5b in ABS at 2.05%, and $2.6b long term debt at 6.43%.  I believe they will optimize their funding mix over the short and intermediate term which will provide a tailwind to EPS not currently captured by the street.  Most obviously, their CD book matures $2.6b in 2011, $10.8b in 2012, $5.9b in 2013, and $6.2b thereafter.  Their current weighted maturity CD offer rate is ~1.33% (you can find the rates on their website) and I am assuming a conservative average maturity of nearly 3 years although I believe they target closer to 2 years.  So their marginal cost of CD funding is1.33%, or 60% lower than their existing CD funding of 3.30% I mentioned earlier.  As such they will refi more than half the CD book by 2012 at less than half the current cost.  Secondly, their savings rates are about 0.30% higher than competitors and they will bring this down as well (move from 1.26% to 1.00%).  In addition, they are growing their savings funding rapidly which will replace their reliance on more expensive ABS funding.  All in by 2012, this mix shift should result in a funding benefit of ~60 bps or $0.40c to EPS which I don’t think is captured by the street. 

    2) Credit costs. 

    Credit costs have been rapidly declining for all card companies and DFS in particular peaked at 9% in q12010 and was only 3.24% in q42011.  Credit costs are coming down for several reasons.  First, although aggregate unemployment is high, the net inflow to unemployment is stable (and has been for some time), and it’s really the event of job loss that drives someone to charge off.  Secondly, industry underwriting has been overly conservative the past several years.  Said another way the loan book is seasoned and most people who will charge off have already charged off.

    It’s commonly understood that in a normal cycle (i.e., not a recession) credit card charge offs will be ~ 5%.  The street is modeling ~3.70% for DFS in 2012 and 4% in 2013, i.e. trying to adhere to the company’s/industry’s comments on the 5% number.  I think however, that we will enter a multi year period where we could meaningfully overshoot credit costs, and the market has not fully realized that and street EPS estimates will have to be revised higher.

    The charge off data is trackable and is reported monthly in the issuer trusts.  30-59 day trust delinquencies have been falling dramatically (driving the initial credit improvement we have seen), but work into charge offs on a lagged basis.  The “roll rate” from the 30-59 day delinquent bucket into the charge off bucket has been pretty consistent over time (in general 75% of those who are 30-59 days late move into the 60-90 day DQ bucket, 85% of 60-89 days DQ roll into 90-120 days DQ, etc….until balances are automatically charged off at 180 days past due.)  On average the conversion rate from 30-59 past due to charge off is ~37% net of recoveries. 

    DFS’ most recent trust data was reported for November at a 3.04% net charge off rate.  That charge off bucket flowed from the cohort of 30-59 DQs in June 2011 which was 72 bps.  Since June 30-59 day DQs have continued to decline to 69 bps as of November.  So for at least the next 6 months charge offs should be in the 2.90% range.  In addition, the last two reported months saw a spike in the roll rate from the 150-180 day DQ bucket to the NCO bucket.  In October and November the roll rate from 150-189 to NCO was 100% and 98% respectively, versus the trailing 12 month average of 92%.  That 92% has been pretty constant over time and I believe it will mean revert.  I think the last two months were driven by exogenous factors (random spike in bankruptcies, estates entering probate etc).  With dollar delinquencies as low as they are even a small move in the roll rate can affect NCOs.  In any event with a 92% roll rate to NCO, NCO’s should be ~ 2.70% for at least the next six months.

    Delinquencies are still falling but at a lower rate than they have been, but it’s worth noting that they fell consistently over the past six months which is usually a seasonally up period for DQs.  In the next six months, the normal seasonal pattern is lower which would argue for NCOs  lower than the 2.70% in 2h12. 

    As far as losses into 2013, the loan book is not growing rapidly (a few % per year), and normal peak charge offs on new originations are usually reached 18-24 months post origination.  Therefore, even as DFS grows its book modestly in 2012, it should be 2014 before that book starts to season and credit costs go up.

    The other way NCOs could go back up would be a massive spike in new unemployment claims.  This is not happening (in fact the opposite) at present, but double dip risk is out there of course.  That said, a sell side firm published a regression of the relationship between the change in new claims versus the change in delinquencies.  Usually delinquencies move up at roughly half the rate of new claims, but it can be volatile in the tails so let's say it's worse and it’s a 1:1 relationship.  So if claims go up 30% from 381k/week currently to 500k/week (implying a huge negative move in the economy and a claim number not seen since the beginning of 2010), DFS NCO’s could go from 2.70% to 3.50% (up 30%).  Barring this substantial retrenchment, I believe we will stay in a prolonged period of abnormally low charge offs.  That said, even if that does happen, the street is at 4% NCOs in 2013! Every 50 bps of NCOs is 35c to EPS.

    So by looking at the current 30-59 day delinquencies, not assuming much improvement from here, but using roll rates which have proven pretty stable over time, it seems DFS should report ~2.70% NCO in 2012 and perhaps 3.00-3.40% in 2013 versus consensus of ~3.60% and 4.00% in both years. 

    3) Revenue suppression.  DFS records contra revenue to account for interest and fee revenue that it will not collect based on its current delinquencies.  On an annualized basis in Q3 this cost them $605m in revenue or $0.70 in EPS.  As delinquencies migrate down this revenue “headwind” will go away as well.  It won’t go to zero (there will always be DQs and charge offs), but if charge offs go to 2.7% as I expect from Q3’s 3.85%, this would be a $180m revenue benefit or $0.21c to EPS in 2012.  Note, I am not baking this in for conservatism, because I think they will have natural yield compression in their book from increased promotional balances (lower APRs), the continued reduction in high rate balances (from delinquent customers that pay the penalty rate but can no longer be repriced due to the CARD ACT), and an increase in customers who pay their balance each month (due to the improving economy).  However, should these other yield compressions prove less onerous, revenue suppression could provide upside. 

    Valuation.  The current consensus estimates are $3.35 in 2012, and $3.37 in 2013.  With lower funding costs, sharply lower losses, and no reserve releases (in fact some build in 2013), DFS should report ~$4.30 in both years.  With these sort of positive earnings revisions I think the stock can conservatively move to the low $30s (8x peak earnings of $4.30).  There will be some recognition that DFS is “overearning” because the market will think charge offs are unsustainably low.  However, given the time it takes new originations to season and that we could be in the < 4% loss range for the next couple years, I think the stock can still work higher.  Even at a 4% NCO ratio, EPS would be $3.50+ which would support a low $30’s price (and in the interim they will be earning more than that).                      

    Conservative balance sheet with significant excess capital.

    DFS’ targeted TCE/TA ratio is 8%, and they are currently at 11.4% which implies $2.3b of excess capital.  DFS generates ~$2.0b of capital per year, so by the end of 2012 they should have ~$4.0b in excess capital (less than the simple addition due to factoring in asset growth) or 31% of the current market cap which will be available for buyback or acquisition.

    Recently the company announced a $1b buyback (8% of shares) which speaks to their prudent capital management.  They bought back 3% of the company in the past two quarters. 

    DFS has significant liquidity with $10b in cash and treasuries (15% of assets) yielding only ~70bps.  They are in the process of investing this liquidity which can be modestly additive to EPS. 

    They are a disciplined acquirer with low risk of a large, lateral acquisition.  There are limited card or student lending assets in the marketplace and they do not want to get into other areas of banking (mortgage, auto). 

    Modest growth.  Card balances are bottoming after 10 quarters of decline from consumer deleveraging.  All the major issuers have shown 2h11 growth.  This is one area where card stocks are differentiating themselves slightly from banks which have pockets of growth (commercial and industrial, card, auto) but have significant run off books (home equity, construction, commercial real estate) that are making their aggregate loan growth zero or negative.  The growth here is modest (low single digits) but it’s still relatively better which should help continue to make these stocks attractive to investors.

    Limited regulatory overhang.  The card companies are one of the areas in financials that don’t have Dodd Frank bill overhangs.  The CARD Act was enacted in 2010 and has been mostly absorbed.  Durbin bill only impacts debit, not credit.  Credit interchange is unlikely to be legislated given how hard it was for Durbin’s debit bill to pass.  The CFPB is mostly focused on disclosure for card companies and will likely attack mortgage before it would seek to do anything else in card.

    Company doesn’t need to exist.  Credit cards are still an attractive business if done properly (highest ROA in financials).  As such DFS could be attractive target for a buyer.  DFS is an odd duckling in that it’s the only monoline card company aside from AXP (COF is another but it’s almost half a bank now).  Also, DFS’s payment network has strategic value because it could be used to internalize one of the major issuers spend volumes away from MA/V (JPM or BAC could own their own network by buying DFS and avoid paying fees to MA/V).  Although I think a major issuer buying DFS for its network is unlikely, I do think at some point when we get more capital clarity over the next 6-12 months, DFS has the potential to be bought as a way for a buyer to expand in the card space.

    Risks -

    Crowded.  Card stocks have been good performers YTD due to the cleanliness of the story relative to other financials, lack of regulatory overhangs, good credit performance, and positive earnings revisions.    

    Timing of revisions.  I’m not sure when the street will really revise to the levels I think they need to.  I can’t make sense of their existing assumptions.  Further revisions should come as we continue get master trust data each month.  Eventually, I think you will get sharp revisions and continued beats that filter into numbers and the stock should follow.

    Catalyst

    Material earnings beats and revisions.  Continued capital deployment.
      Back to top