Gaming FBP, Treasury and the new investors decisions

FBP just filed their latest S-1 filing. This S-1 is not too different but there is this new bit of text:

Although we want to receive gross proceeds of at least $500 million, we may determine to complete an offering for a number of shares of common stock that results in a lower amount of gross proceeds. In that case, we would not be able to fulfill the remaining substantive condition required for us to compel the conversion of the Series G Preferred Stock into common stock, which would mean that our tangible common equity would not benefit from an increase in the number of outstanding shares of common stock resulting from such a conversion and that we will continue to have to accrue dividends payable on the Series G Preferred Stock. In addition, the lower level of proceeds may require us to implement additional de-leveraging strategies to ensure our compliance with the capital plans we submitted to our regulators.

This more or less speaks for itself. They can raise less than $500MM if they think it is in their interest. Whether or not this is in their interest depends on a few things. I will present my game-theory analysis and conclude what I think is in everyone's best interest and therefore what should logically happen.

Right now, the company has the right to compel Treasury to convert their $400MM in preferred equity to 380 million common shares, IF the company raises a cumulative amount of $500 million in common equity elsewhere by the Early Conversion Date which is April 7, 2011. Treasury can convert to 380 million shares any time they wish or can convert any portion.

This is effectively like selling common equity for $1.05 though it does not change total equity or affect the normal capital ratios or anything in the FDIC Consent Order. Since the stock price is currently only $0.30, this right to sell common stock at over a $1 seems like a very valuable asset. But lets looks at the details.

First, we should say that they probably need to raise some capital to satisfy their consent order requirements. However, I do not believe they need to raise the full $500MM. Let's assume that they can get by on less capital and perform other deleveraging transactions to meet the consent order requirements.

In order to simplify the discussion, let's assume the amount they wish to raise is really $300MM. That would cut down the dilution to common shareholders by a large amount. It is easy to see why. New investors will generally want to pay some fixed percentage (a discount naturally) of final tangible book value (TBV). I discussed this in this post. For example, if they desire to pay roughly 45% of final TBV, the two yellow lines in the table show that in the $500MM capital raise scenario, they will pay about $0.40 and the TBV will be $0.91. In the $300MM scenario (with full Treasury conversion), they will pay $0.60 and the TBV will be $1.31.

So the common shareholders come out better with the lower capital raise and less dilution. The Treasury also comes out better since their converted common shares are worth more. There is another difference between the two scenarios. This has to do with the fact that the effective Treasury conversion price of $1.05 (or so) is between the two final TBV values. That means that in the $500MM capital raise scenario, the Treasury conversion is anti-dilutive to book value. In the $300MM scenario, it is dilutive to book value. That means that if they do the $500MM capital raise scenario, the company would not only have the right to compel conversion; they would have the incentive to do so. In the $300MM capital raise scenario, they might not have the right to compel conversion but they also would not really have the incentive to do so since it would be dilutive to TBV. If they do the $300MM raise and Treasury does not convert, then investors would pay $0.70 and TBV would come out to be $1.56.

This is a bit simplistic since I am assuming that investors want to pay 45% of TBV in every scenario. The is probably not the case. But the point is that raising less capital and not having the right to compel conversion is not necessarily a problem. In fact, if they raise $300MM, then Treasury will likely decide to convert on their own since they will probably get more value that way when they exit their investment because final TBV would be $1.31 and, if it trades at TBV within a year or two, the investment would be worth roughly 380*1.31=$498MM which is greater than their $400MM investment with an additional margin of safety.

But what if new investors don't want Treasury to remain unconverted? Perhaps they don't want to have to pay preferred dividends or want the safety of being at the same level in the capital structure as the government. What if they demand to have the right to convert Treasury shares? Does this mean they have to do the $500MM raise? Now you have to think of the this from the Treasury's point of view. What if the company comes to you and asks you to either 1) convert some or all of the Series G shares before the raise or 2) lower the requirement to $300MM so that the company will have the right to compel conversion after the smaller raise.

What would you do if you were Treasury? If you refuse to do either of things you force the company to raise the $500MM and further dilute themselves and also your own position. You make it less likely that you will be able to exit the position without taking a loss on your TARP investment. You have been making a lot lately in public about how you are exciting your TARP investments with no loss or actual gains, showing that you were a good steward of taxpayer's money.

Do you give up anything really by making this concession? Not really. You will be forced to convert anyway since the Company can boost TBV by doing so, and stop the preferred dividends from accruing. If you convert some or all of it now, and allow them to raise only $300MM, it would no doubt give a boost to the stock price and allow them to raise the capital at a better price and maybe even at a higher ratio to final TBV.

So, I think it is logical for the company to ask Treasury for something like this and I think it is logical for Treasury to respond positively to the request. The result would be very positive for the stock. Investors will see that they are going to end up with TBV well above $1 and will probably bid the stock up closer to $0.60, a double from here and a good chance to double again over the next few years.

How to calculate FBP's final tangible book value per share (TBV) and target price

Firstbank (FBP) is going through a capital restructuring and many investors (and analysts) need to be able to value the company based on different outcomes. Usually people try to focus on the final tangible book value per share, TBV, and value the company based on that. One can assign some multiple of TBV as the target price and voila!, you are done.

They have completed the deal with the US Treasury and have successfully completed their preferred conversion. So there are only a few steps left. The major step is the capital raise but Bank of Nova Scotia also has anti-dilution rights. Finally the deal with Treasury also has anti-dilution rights. We need to address all these issues. Let's begin.

First of all, TBV is given by tangible common equity (TCE) divided by total common shares outstanding.

Currently, there are 320 million shares outstanding after the preferred conversion. Tangible common equity is (after Q3 2010), $867MM. So TBV is $867/320 = $2.71/share. The stock is currently at $0.30. Obviously the market expects a lot of dilution.

The company has filed an S-1 with the SEC. According to the S-1 they will raise up to $500MM with another $75MM that can be purchased by the investment banks to cover over allotments if any.

A common misunderstanding is that they are required to raise $500MM. That is not quite correct. They have a consent order to maintain certain capital ratios. They also have a deal with the US Treasury which says: if they raise $500MM, they can compel the conversion of the Series G mandatory convertible shares into common shares. But Treasury may want to convert these anyway and the company might rather NOT convert them. So this agreement with Treasury does not mean that they have to raise $500MM. So lets leave this as a variable which we will call CRA (capital raise amount).

The other very important variable is the capital raise price which we will call CRP. Usually banks raise capital at the market price at the time of pricing or somewhat below it.

There is one more variable that comes into play. That is the market price at the time the capital raise is priced. Let call this MP. The reason that this affects the final TBV is that the deal with Treasury has an anti-dilution provision. If it is not triggered, the Treasury gets 380 million common shares. But if it is triggered, they can get more shares. The provision is triggered if the capital raise price, CRP is less than 90% of the current market price (MP).

The adjustment factor is described in the agreement Section 11. Anti-Dilution Adjustments, paragraph C. Clearly this was written by lawyers but I have translated the legal jargon into math and the number of shares Treasury received if this is triggered (if CRP < 0.9*MP) is (in millions)

N(Treasury) = 380 * (320 + CRA/CRP)/(320 + CRA/(0.9*MP))

Note that if CRP=0.9*MP this factor is 1 and so they get 380 million shares. This assumes they convert all of their Series G shares. If they convert only some fraction then they get this number of common shares multiplied by that same fraction.

Ok, so we are almost done. Lets ignore Scotiabank's anti-dilution provisions for now.

So we have three variables: the capital raise amount (CRA), the capital raise price (CRP) and the market price the day before pricing (MP).

The total number of shares will be (in millions)

N(Total) = 320 + CRA/CRP + 380 * (320 + CRA/CRP)/(320 + CRA/(0.9*MP))

Treasury will (probably) convert $400MM in preferred equity into common equity. The total amount of Tangible common equity will be (in $MM)

TCE = 867 + 400 + CRA

and TBV is TCE/N(Total).

So we need to estimate the variables, CRA, CRP and MP to calculate TBV. Finally, we need to come up with a target price. That is, we need to pick some Price/TBV ratio and multiply TBV by that ratio.

So lets do the standard example.

Since the stock price is currently $0.30, lets assume CRP is the same. Lets assume they raise CRA=$500MM as most people expect and assume that Treasury converts all their Series G shares. So TCE goes to $1767MM. The new investors get CRA/CRP=1667 million shares bringing the total to 1987 million shares excluding Treasury's shares. The number that Treasury gets depends on whether anti-dilution is triggered which depends on whether the market price (MP) is greater than 0.90*CRP or $0.33 in this case. If not, then they get 380 million shares bringing the total share-count to 2367 million shares. In this case TBV is $1767MM/2367 = $0.75. Now, lets say you assign a 1X multiple of TBV for a target price and so your target is also $0.75.

The following tables lists the capital raise price and the resulting final TBV as well as the ratio of the capital raise price to final TBV. The first table id for a $500MM capital raise and the second one is for a smaller $300MM capital raise. The yellow lines highlight the case where the investors demand a price of roughly 45% of final TBV.




Don't worry about Treasury's anti-dilution provision

However, there is one important catch that we should keep in mind when assigning a target price. This has to do with Treasury's anti-dilution. We calculated a TBV and target price of $0.75/share assuming it is not triggered. But what if anti-dilution does get triggered? What if investors over-estimate the pricing at say $0.60 and the final TBV and bid up the market price to $0.80 the day before pricing? Well, then anti-dilution gets triggered and Treasury gets more shares. In this case they get N(Treasury) = 380 * (320 + 500/0.30)/(320 + 500/(0.9*0.80)) = 744 million. So N(Total) is 2731 million and TBV is $0.65. If you apply the same multiple 1X to TBV you come up with a target of $0.65. But hold on! If the stock runs up to $0.80 before the capital raise in this scenario, your target should be $0.80 not $0.64. The general idea is that your target price should be the greater of TBV*(Price/TBV) and MP.

In practice, it is probably best (simplest anyway) just to assume that anti-dilution does not get triggered. That is because the stock price is unlikely to get priced far below the market price. If it does, it will probably be because the stock price runs up a couple of days before the pricing. But if this make a big change in your TBV and target price, the market price will probably exceed your target so it is not a downside risk; it is an upside risk.

Don't worry about Scotia

Finally, there is the issue of Scotia which is also probably best to ignore. The reason is that even if Scotia is really interested in buying into FBP, they would probably not exercise their anti-dilution provisions. They would either buy them out entirely which is also an upside risk or just participate in the public offering. That is because the price in the public offering will almost certainly be lower than the price they would pay by exercising their anti-dilution rights.

If they exercise the anti-dilution rights, they have to pay the same price as each investor paid in each step of the capital restructuring process. They have the right to maintain a 10% stake. That means they can buy 10% of the number of shares that were offered at each step at the same price. Since the capital raise will probably be done at a lower price than the other two steps, the weighted average price will be larger than the capital raise price.

The regular preferred conversion resulted in 228 million new common shares valued at $1.12. So Scotia can buy 22.8 million shares at that same price or $25.5MM. Then they can inject 10% of the capital raise for 10% of the shares that were issued. Finally when Treasury converts, they can inject another $40MM at $0.72 (subject to Treasury anti-dilution). So assuming $500MM capital raise at $0.30 and no Treasury anti-dilution, they could inject $25.5MM+$40MM+$50MM = $115MM and get 227 million shares. That works out to an average of $0.51/share. Why would they do that if they can participate in the public offering at $0.30? Furthermore, if the company knows that Scotia wants to retain their 10% stake but doesn't want to participate in the public offering (for whatever strange reasons), why would they offer $500MM?. If they want only $500MM in new capital, they will offer less and have Scotia inject the rest. For these reasons, it is best to ignore the Scotia anti-dilution rights. If they decide to act on those rights it will likely only be because the stock price has risen way above their effective price putting their options in the money. But then, as before, this is not a downside risk to current investors.

Other adjustments to TCE

There are other adjustments that one might want to do to TCE. For example, you may expect them to have further losses in the next few quarters of say $100MM. I do not but many analysts do. Bain Slack thinks they may have to write-down their Lehman receivable which might be another $50MM. But management has said that they expect to reverse the deferred tax asset valuation allowance which is now $290MM. So even factoring in these losses, they should increase TCE by 290-100-50=$140MM when the DTA valuation allowance gets reversed, probably in 2011. That should add another $0.06 to TBV in the $500MM @ $0.30/share scenario and bring TBV to $0.81/share.

Avoiding or reducing the capital raise: the best scenario

While one should plan on the full $500MM capital raise as the default assumption, there is a reasonable chance that the company will avoid or at least reduce the amount of capital that they need to raise. As I mentioned, they are not required to raise $500MM by Treasury or anyone else. The FDIC consent order, which can be found here , requires that they meet three capital ratio targets "over time". That is, there is no definite time-table to meet those targets. Furthermore, they are currently meeting all three of them with no additional capital. With mild deleveraging, they can probably continue to meet them with no additional capital.

However, there is another requirement having to do with non-perfoming loans.

reduce Special Mention and classified assets to 100% of Tier 1 capital and ALLL, and 75% of Tier 1 capital and ALLL within 6 and 12 months, respectively, from the effective date of this ORDER

The order is dated June 2, 2010 so the 6-month deadline is Dec 2, 2010 and the 12-month deadline is June 2, 2011.

Tier1 capital is currently $1203MM and ALLL is $608MM so the sum is $1811. So they have to have "Special Mention and classified assets" below this level before Dec 2, 2010 or raise enough capital to offset the shortfall. This is the requirement that really determines how much capital they will need to raise.

What is the current value of "Special Mention and classified assets"? Unfortunately, this is a regulatory classification and has not been released in public filings. One can get an estimate of this by looking at the numbers they do report such as Non-performing loans which is $1669MM. Note that this number by itself is already lower than Tier1+ALLL = $1811. But there are probably other Special Mention assets not included in Non-performing loans so this may be more like $2000MM so they may currently have a shortfall of a few hundred million.

On the other hand, NPA's are falling. The Puerto Rico housing stimulus is working very well to increase sales and they may be able to lower NPAs by a few hundred million in the next few months. They have lowered NPAs two quarters in a row now.

They can also sell NPAs. While this may result in further losses, it can only act to improve this "Special mention and classified" requirement. That is because when you sell an asset, you reduce the whole thing from the balance sheet whereas Tier1 capital + ALLL is only reduced by the additional loss than you take. So, for example, lets say they have $700MM (balance due) of NPAs which they are carrying at $500MM. That is, they have charged off $200MM and perhaps they have an additional allowance of $50MM. Lets say they sell them for $350MM or 50 cents on the (balance due) dollar. That removes $500MM from the "Special Mention and classified assets", the left hand side of the equation. But it requires a charge-off of $150MM, a reduction in ALLL of $50MM and a loss to Tier 1 capital of $100MM. The sum Tier1 Capital + ALLL is reduced by the amount of charge-off, $150MM. The improvement in this "Special Mention and classified assets < 100% of Tier 1 + ALLL" condition is $500MM-$150MM = $350MM. That is whenever you sell loans, you improve this condition by whatever proceeds you get. The trade-off is that by taking losses to capital, you lower the other three ratios that have requirements such as Leverage ratio > 8% of which they are very close to now. But as this example demonstrates, if they are short by $350MM, they can raise $350MM of capital or, instead, sell these loans and raise only $100MM to get the Leverage ratio back up to where it was. Also keep in mind that the requirements on the other three capital ratios do not have 6-month deadlines so breaching them by Dec 2, 2010 should not violate the Consent Order.

The point is that selling assets is another option they have that might allow them to raise less capital and still be in compliance with the Consent Order.

One last thing to keep in mind is that the Consent Order provides requirements for the bank only. It is possible that the holding company could borrow money, with either senior debt (probably zero coupon) or convertible debt and inject it into the bank which would put the bank in compliance and give the holding company more time to get back to profitability and if needed, raise capital at a better price in the future (to pay back the debt).

So, I have given different possibilities for how the company might be able to avoid raising capital or at least raise less capital. In these scenarios the TBV could come out anywhere between $1 and $3 depending on exactly how it works out. Lets call it $1.80 on average. Lets say you give this scenario a 30% chance of happening and are assigning a Price/TBV ratio of 1X. If your target in the full-raise scenario was $0.75/share then your weighted average target should be about 0.7*$0.75 + 0.3*$1.80 = $1.06/share. So the chance that they can avoid raising so much capital should add some speculative value to the stock.

Did Puerto Rico experience a housing bubble?

Puerto Rico has been in recession for about five years now. Naturally, the real estate market is very poor with far fewer sales and lower sales prices. Since the US mainland is now recognized to have gone through a dramatic once-in-a-century housing bubble, many people assume that Puerto Rico experienced a similar thing. The trouble is that, until now, publicly available house price indices for Puerto Rico have not existed.

This is beginning to change with the arrival of a house price index for Puerto Rico from the Federal Housing Finance Authority (FHFA). The FHFA was created in 2008 from other federal housing agencies to supervise Fannie Mae and Freddie Mac when they were put into conservatorship. The house price indices that they publish for many different regions are repeat sale indices much like the Case-Shiller Index and so track prices changes in sales of the exact same homes sold at different times. The data comes from records of Fannie Mae and Freddie Mac. The indices for Puerto Rico are described
here and include additional data from the Federal Housing Administration (FHA) and the Federal Home Loan Bank of New York (FHLBNY).

This data for Puerto Rico and other regions is available here .

Without further ado, here is a plot of the house price indices for Puerto Rico and a few other major metropolitan regions: Miami, Los Angeles, Tampa and Chicago. These are the purchase-only, seasonally adjusted numbers. I have normalized them all to 100 at the first quarter of 1995 (where the Puerto Rico data series begins).



So did Puerto Rico experience a housing bubble? Well prices certainly went up. They rose about 86% from the first quarter of 1995 to the peak in the second quarter of 2008, 13 years later. That is a 4.8% annual appreciation. The annual change in CPI over that time is 2.6% so that is a 2.2% real annual appreciation. (The inflation rate in Puerto Rico is actually much higher). The decline from the peak over the last two years is about 7%.

This doesn't really support the viewpoint that Puerto Rico experienced a housing bubble. For example the data for Puerto Rico looks nothing like the dramatic boom-bust seen in the data for Miami, Los Angeles or Tampa. It looks closer to that of Chicago or even milder.

I found some other data here from Freddie Mac. This is a median house price. The chart below shows the median home price appreciation for the US and Puerto Rico over 12 years, up though 2009.



I have taken those annual appreciation numbers and turned them into an "index" by normalizing them to 100 in 1996.



This shows the same thing; that there was no housing bubble in Puerto Rico.

Puerto Rico avoided much of the housing bubble madness for a few reasons. There was no subprime lending in Puerto Rico. Wall Street never took any interest in the Island whose financial system is somewhat old fashioned (now considered a good thing). Loans originated on the Island were either securitized through the federal agencies, Fannie, Freddie, Ginny Mae or FHA or kept on the balance sheet of the banks. In addition, the Island has been in recession since about 2006 so the banks in Puerto Rico were curtailing lending (to some extent at least) well before the bust started in the US two years later. This isn't to say that all is well in Puerto Rico. The five year recession has produced a 10% cumulative decline in GDP which some would call a depression. The levels of bad loans at banks is even higher than on the mainland. But the dynamics were different. The recession in Puerto Rico was not one caused by the exuberance and bust of a real estate bubble. It had other causes that I won't go into here. But the bottom line is that Puerto Rico's housing market did not experience a housing bubble.

Non-performing loan trends

House Price Index and sales for Puerto Rico

A house price index from the FHFA through Q1 2010.



From Estudios Tecnicos through 2008



This chart shows the amazing increase in mortgage originations since the Housing Stimulus began.