TARP Warrants

[stahleyp]

I believe Sanjeev might have been referring to an excel I sent him. I have it in the office, so I'll post it on Monday. It will also give me the chance to update the excel with financial data from the third quarter now that most of the large banks have reported.

 

Regards

 

thanks, damian, it's much appreciated! :)

[damianolive]

Here's the excel. The first tab is based on book value growth and the bank specific tabs are based on a couple of key variables (growth in loans, growth in pretax preprovision profits, charge offs, etc). This obviously over simplifies reality, but the idea is to get a sense of returns under various scenarios - you can play with the variables in blue.

 

The BofA A warrant does not have the strike price reduction feature from dividends, so it needs this adjustment.

 

There's also some historical info on charge offs and aggregate banking system financials that can help as background. Also some info on the experiences from the early '90s on a few banks (note Wells Fargo actually recorded zero provisions in '95).

 

Of interest, in the last 3 majors charge off cycles (mid seventies, early nineties, early 2000s) it took exactly 4 years to go from peak to trough. If we assume that the peak was in the 4Q/2009 (the tab called 2010 shows this for the big banks ex Citi), then we're in for a slow and steady process of charge off reductions. This is at odds with what Walls Street is projecting for some major banks (e.g. COF); this cycle might be longer, but it's hard to see how charge offs will simply stay where they are, which is what they assume.

 

Finally, I did not have time to update data from the 3Q, but this can be easily done in tab 2010.

[RRJ]

Thanks very much for posting this.  It's obvious a lot of work went into this.  Really lays out the comparison nicely.

[claphands22]

 

The BofA A warrant does not have the strike price reduction feature from dividends, so it needs this adjustment.

 

 

 

These warrant prospectuses are hard to read and you've done way more research on these warrants than I have. Yet, I thought the BofA A warrants also had the strike price reduction from dividends.  I'm looking at the prospectus (http://www.sec.gov/Archives/edgar/data/70858/000119312510044940/d424b7.htm#suptx70043_6) page S-28 -  S-30 under "Adjustments to Warrants"

 

Any kind of clarification will help.

[Uccmal]

Its in there claphands...  If BAC raises the dividend above 0.01 (the present dividend) per Q it starts reducing the

ACB of the warrants.  I think Damienolive means that his table does have the feature.  Anyways, a great layout Damien.  Thanks. 

[dcollon]

Very nice spreadsheet.  Thank you for sharing it with the board.

 

 

[vinod1]

Here's the excel. The first tab is based on book value growth and the bank specific tabs are based on a couple of key variables (growth in loans, growth in pretax preprovision profits, charge offs, etc). This obviously over simplifies reality, but the idea is to get a sense of returns under various scenarios - you can play with the variables in blue.

 

The BofA A warrant does not have the strike price reduction feature from dividends, so it needs this adjustment.

 

There's also some historical info on charge offs and aggregate banking system financials that can help as background. Also some info on the experiences from the early '90s on a few banks (note Wells Fargo actually recorded zero provisions in '95).

 

Of interest, in the last 3 majors charge off cycles (mid seventies, early nineties, early 2000s) it took exactly 4 years to go from peak to trough. If we assume that the peak was in the 4Q/2009 (the tab called 2010 shows this for the big banks ex Citi), then we're in for a slow and steady process of charge off reductions. This is at odds with what Walls Street is projecting for some major banks (e.g. COF); this cycle might be longer, but it's hard to see how charge offs will simply stay where they are, which is what they assume.

 

Finally, I did not have time to update data from the 3Q, but this can be easily done in tab 2010.

 

damianolive,

 

Outstanding effort. Thanks for sharing.

 

Vinod

[claphands22]

Its in there claphands...  If BAC raises the dividend above 0.01 (the present dividend) per Q it starts reducing the

ACB of the warrants.  I think Damienolive means that his table does have the feature.  Anyways, a great layout Damien.  Thanks.   

 

Thanks for clarifying.

[stahleyp]

Here's the excel. The first tab is based on book value growth and the bank specific tabs are based on a couple of key variables (growth in loans, growth in pretax preprovision profits, charge offs, etc). This obviously over simplifies reality, but the idea is to get a sense of returns under various scenarios - you can play with the variables in blue.

 

The BofA A warrant does not have the strike price reduction feature from dividends, so it needs this adjustment.

 

There's also some historical info on charge offs and aggregate banking system financials that can help as background. Also some info on the experiences from the early '90s on a few banks (note Wells Fargo actually recorded zero provisions in '95).

 

Of interest, in the last 3 majors charge off cycles (mid seventies, early nineties, early 2000s) it took exactly 4 years to go from peak to trough. If we assume that the peak was in the 4Q/2009 (the tab called 2010 shows this for the big banks ex Citi), then we're in for a slow and steady process of charge off reductions. This is at odds with what Walls Street is projecting for some major banks (e.g. COF); this cycle might be longer, but it's hard to see how charge offs will simply stay where they are, which is what they assume.

 

Finally, I did not have time to update data from the 3Q, but this can be easily done in tab 2010.

 

 

you're the best! thanks!!

[redskin]

If BAC pays out the dividends that are assumed in the spreadsheet (approximately $50B) the strike price on the 'A' warrants will be reduced by $5.  It will go from 13.30 to 8.30.  This increases the IRR substantially.  I think you could also expect that there will be some share buybacks which should increase the EPS.  I wouldn't be surprised to see BAC returning 50% of their earnings in 2-3 years through dividends and share buybacks depending on the level of the stock price.  Maybe 25% divys and 25% share buybacks?

[ValueBuff]

If BAC pays out the dividends that are assumed in the spreadsheet (approximately $50B) the strike price on the 'A' warrants will be reduced by $5.  It will go from 13.30 to 8.30.  This increases the IRR substantially.  I think you could also expect that there will be some share buybacks which should increase the EPS.  I wouldn't be surprised to see BAC returning 50% of their earnings in 2-3 years through dividends and share buybacks depending on the level of the stock price.  Maybe 25% divys and 25% share buybacks?

 

You beat me to it.  This is exactly what I was going to post.  With the 50% payout ratio, which imo could be 10% higher even, your strike falls to 7.78$ per warrant.  Keep in mind the warrant is tax deffered, whereas the dividend might be taxable as you receive it.

 

Did you use the book value given by B of A in their last earnings?  BAC has almost shed 90 Billion in marketcap since Feb '10.  Bad news priced in?

[twacowfca]

There's a non obvious downside to a reduction of strike price in lieu of dividend.  The value of a strike price reduction won't compound from that point in time until the warrant expires.   :(

[redskin]

There's a non obvious downside to a reduction of strike price in lieu of dividend.  The value of a strike price reduction won't compound from that point in time until the warrant expires.   :(

 

Not sure what you mean by this?

[Uccmal]

There's a non obvious downside to a reduction of strike price in lieu of dividend.  The value of a strike price reduction won't compound from that point in time until the warrant expires.  :(

 

That is true and well thought out twacowfca.  It will be a straight dollar per dollar conversion.  Perhaps what you lose on the dividend compounding you gain on the tax deferral?  You get to keep the money invested rather than paying tax on it and paying tax on the compunded amount as well. 

[twacowfca]

There's a non obvious downside to a reduction of strike price in lieu of dividend.  The value of a strike price reduction won't compound from that point in time until the warrant expires.   :(

 

That is true and well thought out twacowfca.  It will be a straight dollar per dollar conversion.  Perhaps what you lose on the dividend compounding you gain on the tax deferral?  You get to keep the money invested rather than paying tax on it and paying tax on the compunded amount as well. 

 

 

Yes.  However,  The advantage would be definitely in favor of receiving a dividend in a tax deferred or tax free account.  :)

[redskin]

There's a non obvious downside to a reduction of strike price in lieu of dividend.  The value of a strike price reduction won't compound from that point in time until the warrant expires.   :(

 

Not sure what you mean by this?

 

I think I understand what you are saying.  By paying a dividend it will reduce their retained earnings?  I understand the point, but if they are making 20-30B per year, I don't think they will be able to efficiently reinvest all of it.  Will work out better to pay out.  I personally think the best option for shareholders and holders of the warrants will be for them to buy back shares (if they are allowed) while prices are anywhere near where they are now.  When they get the ok from regulators I hope they aggressively buy back shares rather than hike the dividend initially.

[Parsad]

Thanks for posting that Damian!  I did not want to put the spreadsheet out unless you felt it was ok.  The tweeks I made were to make the ROE numbers and payouts much more conservative.  That will give you a better margin of safety.  I also made a few other additions based on capital, etc, that unfortunately I cannot share...they are now proprietary!  ;D  Cheers!

[twacowfca]

There's a non obvious downside to a reduction of strike price in lieu of dividend.  The value of a strike price reduction won't compound from that point in time until the warrant expires.   :(

 

Not sure what you mean by this?

 

I think I understand what you are saying.  By paying a dividend it will reduce their retained earnings?  I understand the point, but if they are making 20-30B per year, I don't think they will be able to efficiently reinvest all of it.  Will work out better to pay out.  I personally think the best option for shareholders and holders of the warrants will be for them to buy back shares (if they are allowed) while prices are anywhere near where they are now.  When they get the ok from regulators I hope they aggressively buy back shares rather than hike the dividend initially.

 

 

That's a good point.  Buybacks at an attractive stock price, below IV, may be more tax efficient than returning value to shareholders through dividends, but that wasn't what I was trying to convey.  Most warrants don't allow the warrant holder to receive dividends when they are distributed, although some warrants have that nice feature.  

 

In a situation such as the TARP warrants, the value received by the warrant holder, a reduction in strike price corresponding to the amount of the dividend paid above a certain threshold, to compensate for the loss in BV in the underlying security after payment of a dividend, will not be as great as the value of the dividend above the threshold would be to the shareholder in a tax free account.  The shareholder would be able to reinvest the dividend and earn interest that would compound.  However, the holder of the warrant who received a reduction in strike price would not be able to do this.  The value of the corresponding reduction would not compound through the remaining life of the warrant.  This will be a considerable difference for a warrant that does not expire for some  years after a dividend is paid.  :)

[ericd1]

Question for all...Let's assume the numbers on some of these warrants look good for a potential return greater then 20% with a decent margin of safety. That's a large enough compound return to make me interested.  But because they are warrants, they might not trade in-line with their underlying stock until close to the expiration date. It could be that the required holding period would be until expiration in 2018-2019, or about 7, or 8 years. That seems like a long time to me to find out if my initial analysis was correct.  Unless you are Buffett with a holding period of forever, I'm not sure these would have a place in my portfolio, even though they look attractive.

 

So after the rambling, my question is--How long are you willing to hold an under-valued position to see if it pans out?

 

If I were a good as Buffett with my picks I'd hold them forever, but I'm not...so my answer to the question is 2 to 3 years if I expect a double, or triple...and 12-18 months if the expected return is lower and the underlying business doesn't seem to be changing...

[twacowfca]

Question for all...Let's assume the numbers on some of these warrants look good for a potential return greater then 20% with a decent margin of safety. That's a large enough compound return to make me interested.  But because they are warrants, they might not trade in-line with their underlying stock until close to the expiration date. It could be that the required holding period would be until expiration in 2018-2019, or about 7, or 8 years. That seems like a long time to me to find out if my initial analysis was correct.  Unless you are Buffett with a holding period of forever, I'm not sure these would have a place in my portfolio, even though they look attractive.

 

So after the rambling, my question is--How long are you willing to hold an under-valued position to see if it pans out?

 

If I were a good as Buffett with my picks I'd hold them forever, but I'm not...so my answer to the question is 2 to 3 years if I expect a double, or triple...and 12-18 months if the expected return is lower and the underlying business doesn't seem to be changing...

 

 

Lack of liquidity often leads to a discount in valuation for lightly traded options and warrants, but this should not be a problem with most TARP warrants because of the size of the issues.  Long term leaps or warrants are often attractive compared to owning common stock, but only if there is good reason to buy the stock itself.  It's generally not a good idea to buy warrants if one would not buy the stock.  :)

[jasonw1]

With the current share vs warrant price, I think common is a better deal than warrant, with less risk.

 

I used the spreadsheet and adjusted price to mid point of PE=10 and PB=1, and got the share price of 32, which I feel it's comfortable and not too aggressive, then return of common and warrant A are about the same, both at 3X of the initial price without counting dividends, while warrant B is negative.

 

 

 

[Ross812]

Just a thought.  Aren't warrants issued by the company?  I thought the reason WFC bought a portion of the warrants back was to prevent dilution.  Is it important to include the dilutive effect from warrant exercise into the models being built?  I thought I would add (if not already shares) Chou bought 1.2 million BAC A warrants prior to his 6/30 report at 5.96 a piece. 1.9% of his portfolio.      

[jasonw1]

For BAC, there are ~150 million A warrant and ~120 million B warrant, yes they will be dilutative but fairly small considering current 10 billion shares of common stock

 

Not sure the exact time when Chou bought it A warrant but BAC common has been trading at $15-19 range for the first half of this year, the price now is $11. I also wonder how Chou bought A warrant at <$6, as its starting bid was $7 and never traded below $6 until in the last couple of weeks.

[shalab]

He paid 8.54 per warrant. ( 1.2 million warrants at 10.245 million dollars ). You can now buy it at 30% discount to Chou.

 

 

[twacowfca]

Just a thought.  Aren't warrants issued by the company?  I thought the reason WFC bought a portion of the warrants back was to prevent dilution.  Is it important to include the dilutive effect from warrant exercise into the models being built?  I thought I would add (if not already shares) Chou bought 1.2 million BAC A warrants prior to his 6/30 report at 5.96 a piece. 1.9% of his portfolio.      

 

BV/SH is often a good method for valuing financial companies, but only if the BV is solid.  Fully diluted BV/SH is better.  This takes into account the potential dilution and the money received for the warrants if exercised. Right now, there would be little or no negative effect on BV from exercise of warrants because all TARP warrants are out of the money. Exercise of warrants currently would not dilute BV/SH significantly in many cases because most banks are not selling at a premium to stated BV.  

 

There may have been a charge against BV when warrants were issued to the government, but with all the funny money accounting allowed for banks, this should be verified.  Example: about half the recent earnings reported by big banks was from adjustments to their loan loss reserves.  Try "taking that to the bank"!