BAC-WT - Bank of America Warrants

[ERICOPOLY]

So is $500m still the normalized quarterly LAS cost level even after selling such a huge chunk of the servicing portfolio?

 

[ERICOPOLY]

The messy Countrywide-originated loan files in the servicing portfolio run off and gradually get replaced with BAC in-house retail originated mortgages that are properly documented, files are fully in order, etc...

 

Thus, in the future the average mortgage that goes bad will be cheaper to foreclose on than the average mortgage in today's pipeline.

 

Part of the explanation maybe.

[racemize]

 

One thing that is a head scratcher to me - so BAC sold off the delinquent mortgages for a profit.  And they say this leads to a large decrease in costs.  So this is a win/win/win -- selling the MSRs increases B3 capital, it's a profitable sale, and it's a big reduction in expenses. 

 

The buyer is either a sucker or somehow knows how to foreclose a lot more efficiently than BAC. 

 

 

I agree--perhaps they had to sell some that were still going to be profitable to get them to take the nasty ones.  In the call, I remember them implying or saying there would be two batches of transfers, with the second batch having the larger impact on the 60+ days del.

[xazp]

That is a good point.

 

Though, the obvious way to handle the foreclosure settlement would be to forgive/modify/short-sale the loans on the foreclosures that were most costly to process, and/or where they had the worst documentation. 

 

I read somewhere that a portion of these costs are fixed.  I could guess: specialty computer systems, phone banks, lawyers, management - that charge the same whether they are processing 1MM or 500,000 mortgages. 

 

FWIW the countrywide data is interesting.  See pages 36 and 37 of the presentation. 

 

1)  In HELOCs, less than 10% of the countrywide portfolio is 180+ past due.  This is interesting because they've got $8.5Bn of loans remaining and $2.4Bn of reserves.  So, even if that 10%, or double that to 20% - are total losses, that wouldn't equal the reserves.

 

2)  In mortgages, the charge-off rates for Countrywide mortgages are lower than the chargeoffs for BAC excl. countrywide. 

 

This suggests to me that either through the foreclosure settlement or other factors, the Countrywide port has really gotten rid of the bad junk. 

 

 

 

 

 

 

The messy Countrywide-originated loan files in the servicing portfolio run off and gradually get replaced with BAC in-house retail originated mortgages that are properly documented, files are fully in order, etc...

 

Thus, in the future the average mortgage that goes bad will be cheaper to foreclose on than the average mortgage in today's pipeline.

 

Part of the explanation maybe.

[finetrader]

Thoughts on where things get better to the 13% to 15% level from here?

 

1)In the CC they mentioned that the effective tax rate is 30%

2)oh say can you say stock buyback

[vinod1]

 

2)  In mortgages, the charge-off rates for Countrywide mortgages are lower than the chargeoffs for BAC excl. countrywide. 

 

This suggests to me that either through the foreclosure settlement or other factors, the Countrywide port has really gotten rid of the bad junk. 

 

 

My understanding is that this is due to PCI accounting. PCI loans do not have a charge off associated with them as they are marked down to a price that incorporates the loan losses estimates over the life of the loan.

 

Vinod

[xazp]

I may be mistaken, but my understanding is:

 

With a non-PCI loan, you add loan-loss reserves which reflect your anticipated charge offs for some future period (say, the next year or two).  When the charge-offs happen, they subtract from the loan-loss reserves.  The company keeps adjusting their LLRs (up or down) reflecting their view of the future, and the charge-offs happen when someone becomes late and/or defaults on their mortgage. 

 

With a PCI loan, they anticipate the life-time charge-offs during the acquisition process.  They are often aggressive about these estimates because there are tax benefits to doing so.  The company doesn't adjust the PCI reserves, so charge-offs keep whittling the number down. 

 

At least with my understanding, the reserves don't change, but charge-offs represent the actual losses (meaning actual delinquencies, foreclosures, etc) being experienced. 

 

 

 

 

 

2)  In mortgages, the charge-off rates for Countrywide mortgages are lower than the chargeoffs for BAC excl. countrywide. 

 

This suggests to me that either through the foreclosure settlement or other factors, the Countrywide port has really gotten rid of the bad junk. 

 

 

My understanding is that this is due to PCI accounting. PCI loans do not have a charge off associated with them as they are marked down to a price that incorporates the loan losses estimates over the life of the loan.

 

Vinod

[racemize]

Wells Fargo on pci accounting.

 

http://secfilings.nyse.com/filing.php?ipage=7345009&DSEQ=&SEQ=35&SQDESC=SECTION_PAGE

[ERICOPOLY]

 

One thing that is a head scratcher to me - so BAC sold off the delinquent mortgages for a profit.  And they say this leads to a large decrease in costs.  So this is a win/win/win -- selling the MSRs increases B3 capital, it's a profitable sale, and it's a big reduction in expenses. 

 

The buyer is either a sucker or somehow knows how to foreclose a lot more efficiently than BAC. 

 

 

I agree--perhaps they had to sell some that were still going to be profitable to get them to take the nasty ones.  In the call, I remember them implying or saying there would be two batches of transfers, with the second batch having the larger impact on the 60+ days del.

 

First, they transfer 1,768,000 current loan MSRs early on

Second, they later transfer 232,000 non-performing loan MSRs.

Meanwhile, their fees from servicing these loans offset their expenses such that the delta is expected to be "negligible for the year"

 

That suggests that overall this transfer is going to cost us profits.  So it's no wonder there was an interested buyer, it represents a profitable chunk of business that they sold.

 

To clarify in differnet words, if you collect a ton of fees for only a short while, and yet this substantially offsets a ton of expenses that you incur for a long while, then one would tend to believe that BAC would have made an overall profit on these serviced loans were they to all be transferred at the exact same time.

 

 

 

Quoting from the CC transcript:

 

We've referenced our January 7 announcement of agreements to sell MSRs totaling $306 billion aggregate unpaid principal balance. This represents 2 million loans of which 232,000 are 60 plus day delinquent. The transfers of these servicing rights are scheduled to occur in stages over the course of 2013 with the delinquent loans scheduled to be transferred after the current loans. Currently, we recognized approximately $200 million in servicing fees per quarter associated with these loans, which is expected to decrease throughout the year, as we actually transferred the servicing. However, the impact on earnings from lower revenue is expected to be negligible for the year, as we expect expenses to also decrease as we transfer the servicing, especially, the 60 plus day delinquent loans.

[ERICOPOLY]

Unfortunately, BAC has provided us with a normalized run rate of $500m per quarter LAS expenses, but never gave us a normalized run rate of quarterly LAS revenue.

 

So for example I'm expecting a $2.5b quarterly decline in LAS expenses, but this recent sale reduced quarterly revenue by $200m.  Do they have plans to grow their total number of serviced loans from here, or further shrink them?

[QLEAP]

I may be mistaken, but my understanding is:

 

With a non-PCI loan, you add loan-loss reserves which reflect your anticipated charge offs for some future period (say, the next year or two).  When the charge-offs happen, they subtract from the loan-loss reserves.  The company keeps adjusting their LLRs (up or down) reflecting their view of the future, and the charge-offs happen when someone becomes late and/or defaults on their mortgage. 

 

With a PCI loan, they anticipate the life-time charge-offs during the acquisition process.  They are often aggressive about these estimates because there are tax benefits to doing so.  The company doesn't adjust the PCI reserves, so charge-offs keep whittling the number down. 

 

At least with my understanding, the reserves don't change, but charge-offs represent the actual losses (meaning actual delinquencies, foreclosures, etc) being experienced. 

 

 

 

 

 

2)  In mortgages, the charge-off rates for Countrywide mortgages are lower than the chargeoffs for BAC excl. countrywide. 

 

This suggests to me that either through the foreclosure settlement or other factors, the Countrywide port has really gotten rid of the bad junk. 

 

 

My understanding is that this is due to PCI accounting. PCI loans do not have a charge off associated with them as they are marked down to a price that incorporates the loan losses estimates over the life of the loan.

 

Vinod

 

 

Ed Najarian - ISI Group

And then I guess my second question is just fairly technical. But when I look at your -- what you’ve outlined in

terms of reserve recapture, it looks like about $900 million in terms of loan loss reserve, a $2.2 billion provision

and $3.1 billion charge-offs. But it looks like the loan loss reserve itself dropped by about $2 billion from the

third quarter. Can you reconcile that for me?

Bank of America's CEO Discusses Q4 2012 Results - Earnings Call Tra... http://seekingalpha.com/article/1118111-bank-of-america-s-ceo-discuss...

Bruce Thompson

Yeah, the reason is it dropped by that amount is that and you saw it in the third as well as the fourth quarter that

some of the DOJ AG settlement modification and other things, that is as you dispose, get repaid or write off the

purchase credit impaired portfolio, it reduces your loan loss reserve.

 

Ed Najarian - ISI Group

And then that’s not coming through the charge-off line?

Bruce Thompson

That’s correct.

[vinod1]

I may be mistaken, but my understanding is:

 

With a non-PCI loan, you add loan-loss reserves which reflect your anticipated charge offs for some future period (say, the next year or two).  When the charge-offs happen, they subtract from the loan-loss reserves.  The company keeps adjusting their LLRs (up or down) reflecting their view of the future, and the charge-offs happen when someone becomes late and/or defaults on their mortgage. 

 

With a PCI loan, they anticipate the life-time charge-offs during the acquisition process.  They are often aggressive about these estimates because there are tax benefits to doing so.  The company doesn't adjust the PCI reserves, so charge-offs keep whittling the number down. 

 

At least with my understanding, the reserves don't change, but charge-offs represent the actual losses (meaning actual delinquencies, foreclosures, etc) being experienced. 

 

 

2)  In mortgages, the charge-off rates for Countrywide mortgages are lower than the chargeoffs for BAC excl. countrywide. 

 

This suggests to me that either through the foreclosure settlement or other factors, the Countrywide port has really gotten rid of the bad junk. 

 

 

My understanding is that this is due to PCI accounting. PCI loans do not have a charge off associated with them as they are marked down to a price that incorporates the loan losses estimates over the life of the loan.

 

Vinod

 

Agree completely with non PCI loan accounting.

 

For PCI, there are no reserves at the time the loan is acquired. We may be saying the same thing but let me give an example of how I think this works.

 

Say a loan has a principal balance of $10k, 5 year term, interest rate of 10% and one year into the loan term it did not pay the required interest. When a bank purchases this loan they mark it down to a loan value of say $6000 when they expect to say recovery of $2000 annually over the next 4 years for a total undiscounted recovery of $8000. They would mark the loan as a PCI loan for $6000, and accrue interest at 12.6%. No reserves are established at this stage. They would record an accretable yield of $2000 but it is not on BS.

 

One year down, say they expect the recovery of less than $2000 annually, then they are required to establish a loan loss reserve from that point on. The accounting at this stage is very similar to non-PCI loans. The writeoffs and reduction in LLR mentioned QLEAP would be perfectly consistent with this.

 

Vinod 

[vinod1]

My point is that we cannot compare the charge off rates for countrywide when they are PCI loans with original BAC loans as countrywide PCI loans  have been marked to fair value.

 

Vinod

[vinod1]

 

First, they transfer 1,768,000 current loan MSRs early on

Second, they later transfer 232,000 non-performing loan MSRs.

Meanwhile, their fees from servicing these loans offset their expenses such that the delta is expected to be "negligible for the year"

 

That suggests that overall this transfer is going to cost us profits.  So it's no wonder there was an interested buyer, it represents a profitable chunk of business that they sold.

 

To clarify in differnet words, if you collect a ton of fees for only a short while, and yet this substantially offsets a ton of expenses that you incur for a long while, then one would tend to believe that BAC would have made an overall profit on these serviced loans were they to all be transferred at the exact same time.

 

 

This is exactly what Plan has pointed out to me a little back in the thread. MSR's at this time seem to be generating pretty good IRR.

 

I think BAC wants to get rid of servicing non-core customers, even if it slightly reduces their profits. Sort of like getting rid of ILFC by AIG.

 

Vinod

[xazp]

OK a few collated replies:

- Vinod, etc: I think we're pretty much on the same page now (that is, you've raised some point that I had missed before). 

 

- Eric:  you raise a good point about LAS revenues, and it likely means I've overestimated the impact of the $10Bn cost-cutting number that we both talk about. 

 

However, some important points. 

 

MSRs are a special case in capital allocation for BAC.  They don't contribute presently to B3 capital.  So selling $1Bn of MSRs immediately generates $1Bn of B3 capital.  And $1Bn in B3 capital = $1Bn in buybacks, or reinvestment in other parts of the business.    Essentially $1Bn in capital, however it is used, is preferable to $0 in capital and whatever profits the MSR produces. 

 

My assumption is they're not selling random MSRs, but rather (for example) MSRs that are attached to a legacy computer system/platform, or which are poorly documented.  So my hope is that this process leads to a drop in some fixed costs (i.e. whatever it costs to run and maintain some legacy systems).  The sensible process to me would be to get rid of every MSR that is on a legacy system, shut down those systems.  Then, if they want to expand their MSR business in the future they can do that by issuing their own mortgages and putting it on whatever the new MSR system is going forawrd. 

 

 

 

 

 

 

 

 

I may be mistaken, but my understanding is:

 

With a non-PCI loan, you add loan-loss reserves which reflect your anticipated charge offs for some future period (say, the next year or two).  When the charge-offs happen, they subtract from the loan-loss reserves.  The company keeps adjusting their LLRs (up or down) reflecting their view of the future, and the charge-offs happen when someone becomes late and/or defaults on their mortgage. 

 

With a PCI loan, they anticipate the life-time charge-offs during the acquisition process.  They are often aggressive about these estimates because there are tax benefits to doing so.  The company doesn't adjust the PCI reserves, so charge-offs keep whittling the number down. 

 

At least with my understanding, the reserves don't change, but charge-offs represent the actual losses (meaning actual delinquencies, foreclosures, etc) being experienced. 

 

 

2)  In mortgages, the charge-off rates for Countrywide mortgages are lower than the chargeoffs for BAC excl. countrywide. 

 

This suggests to me that either through the foreclosure settlement or other factors, the Countrywide port has really gotten rid of the bad junk. 

 

 

My understanding is that this is due to PCI accounting. PCI loans do not have a charge off associated with them as they are marked down to a price that incorporates the loan losses estimates over the life of the loan.

 

Vinod

 

Agree completely with non PCI loan accounting.

 

For PCI, there are no reserves at the time the loan is acquired. We may be saying the same thing but let me give an example of how I think this works.

 

Say a loan has a principal balance of $10k, 5 year term, interest rate of 10% and one year into the loan term it did not pay the required interest. When a bank purchases this loan they mark it down to a loan value of say $6000 when they expect to say recovery of $2000 annually over the next 4 years for a total undiscounted recovery of $8000. They would mark the loan as a PCI loan for $6000, and accrue interest at 12.6%. No reserves are established at this stage. They would record an accretable yield of $2000 but it is not on BS.

 

One year down, say they expect the recovery of less than $2000 annually, then they are required to establish a loan loss reserve from that point on. The accounting at this stage is very similar to non-PCI loans. The writeoffs and reduction in LLR mentioned QLEAP would be perfectly consistent with this.

 

Vinod

[Liberty]

http://www.nytimes.com/2013/01/20/business/a-fed-voice-asking-to-cut-megabanks-down-to-size.html

 

 

[txlaw]

Can I just say that I really enjoy this thread?

 

The discussion and analysis here is phenomenal.  Thanks to all you guys who post so much on BAC.

[damianolive]

Here's an attempt to put some numbers into the servicing costs and income question. Bottom line it looks like about $0.58 EPS benefit from LAS expense reduction and around $0.08 decline in servicing income, for a net benefit of $0.50.

BAC_LAS_Analysis.xlsx

[Uccmal]

Can I just say that I really enjoy this thread?

 

The discussion and analysis here is phenomenal.  Thanks to all you guys who post so much on BAC.

 

Agreed, this is what makes this board great.  I have learned alot in the last few pages.  A.

[nkp007]

Can I just say that I really enjoy this thread?

 

The discussion and analysis here is phenomenal.  Thanks to all you guys who post so much on BAC.

 

Agreed, this is what makes this board great.  I have learned alot in the last few pages.  A.

 

Someone should make this entire thread into a book. Given what we've made and will make, I'm sure a few of us would be glad to put it on our coffee tables.

[xazp]

I wanted to ask the board about a question that has puzzled me for quite some time.  And that is, where are the severance costs?  Surely you can not cut $18 billion of costs without paying a few dollars in severance. 

 

Compare C to BAC's earnings.  C reported ~$1Bn of "repositioning" in Q4, and ~$1.35Bn for 2012.  So the analysts dutifully add in ~$1Bn of pre-tax to come up with C's adjusted Q4 eps.

 

Unless I am completely missing something, BAC has never reported a severance or "repositioning" number for each quarter or year.  So, the analysts pretty much assume it is zero and do not adjust the eps. 

 

But if anything, unless BAC has some magical formula - the severance number for BAC has to be higher.  They're cutting a lot more in expenses. 

 

Is there any way to guess how much BAC has incurred in severance?  Because you could also intuit that everyone is assuming a depressed level of "current" earnings - unless I have missed something.  Seems like it could be material - hard to believe for example that there was less than $1.3Bn (C's level) of severance in the last year. 

 

 

 

[berkshiremystery]

Can I just say that I really enjoy this thread?

 

The discussion and analysis here is phenomenal.  Thanks to all you guys who post so much on BAC.

 

Agreed, this is what makes this board great.  I have learned alot in the last few pages.  A.

 

Someone should make this entire thread into a book. Given what we've made and will make, I'm sure a few of us would be glad to put it on our coffee tables.

 

I feel like in the good old Fairfax days, while the market would feel uncertain, this board would calculate the CDS gains and it's gains on BV. There should be no fear in BofA, if we can assume that the hidden gravity force field is around PTPP $1.60 for fiscal year 2013, even if the reported future actual numbers might be around ~$1 EPS at yearend 2013. This hidden gravity field should slowly pull the pig through the python northward. The last year and this are probably like 1988/89 as it grew 100% and 50% in the next year (quoting Tom Brown). So it might be not unreasonable to see some nice Christmas 2013 valuation for the common in the ranges $16-17.5, and for the a-warrants above $7.5. But someone should be aware that it might become dangerous to hold the shorter maturity Jan 2014 Leaps cycle, because in a sideway market there would be definitely some time-value decay.

[rkbabang]

http://www.nytimes.com/2013/01/20/business/a-fed-voice-asking-to-cut-megabanks-down-to-size.html

 

 

What is your opinion of this?  Have you factored in the possibility of the government doing something drastic, completely changing the company and its potential, into your BAC investment thesis?  I haven't.  That article makes me a bit nervous.

 

 

 

[berkshiremystery]

http://www.nytimes.com/2013/01/20/business/a-fed-voice-asking-to-cut-megabanks-down-to-size.html

 

 

What is your opinion of this?  Have you factored in the possibility of the government doing something drastic, completely changing the company and its potential, into your BAC investment thesis?  I haven't.  That article makes me a bit nervous.

 

There are roughly 5,600 commercial banking institutions in the country, Mr. Fisher noted. Some 5,500 of them are community banks. While these organizations account for 98.6 percent of all banks, they hold only 12 percent of total industry assets. They are routinely allowed to fail if they get into trouble. Few of them did during the crisis.

 

Contrast these figures with those of the nation’s 12 largest banks, whose assets range from $250 billion to $2.3 trillion. They account for 0.2 percent of all banks but hold 69 percent of industry assets. These are the banks that enjoy all the perquisites of the federal safety net: significantly lower borrowing costs and a taxpayer backstop, for example.

 

-----

 

Every industry has it's own type of 80/20 principle, or maybe even odder like 90/10, 99/5, or this extreme 69/0.2 like mentioned in the article. I have no opinion if lawmakers might be able to change rules against the lobbying powers, but certainly it might be good. But in the end capitalism depends on the gravity of these 80/20 principles, and one way or another there will be big banks or other companies able to do something stupid. Lawmakers would only change a small needle in a haystack, maybe get it from 69/0.2 to about 50/0.9 But it's silly to assume that we ever get to a 100/100 principle, where 1% banks have 1% of the assets, i.e. 10/10 and so on. I guaranty it, it will never, never ever happen, because not even the universe and it's galaxies are evenly spread, and such a world economc order would have no incentives. I bet even in China are mega banks.

[berkshiremystery]

Bank Of America Will Likely Double Again Within 3 Years

2013-01-21 SeekingAlpha.com

 

http://seekingalpha.com/article/1122441-bank-of-america-will-likely-double-again-within-3-years?source=google_news