Thursday, November 29, 2007

E*Trade: There's no sense in your risking yourself on my account

Ah it seems like only yesterday, but it was a whole two weeks ago when E*Trade appears to be on the verge of bankruptcy. Now Citadel has infused them with $2.5 billion, pulling them back from the brink.

I think this is note worthy on a variety of fronts. But most important is that it is possible to make a reasonable investment in a company with negative net worth. Many commentators, including some of AI's readers have questioned why anyone would put equity capital into various subprime-tainted companies. And I get their logic. Why put cash into a company laden with losses, especially if the losses are so great as to create negative net worth? It would sure seem like throwing away good money after bad.

Its relatively simple to build a model showing why sometimes investing in a negative net-worth situation makes sense. This will be important for any number of companies, from Washington Mutual to Ambac/MBIA/FGIC to Citigroup and Freddie Mac.

Bear in mind that I have no view of the E*Trade deal, since that is a company I don't follow at all, and therefore have no idea what may or may not make sense in their particular case. But the model I will show here could apply to anyone who has suffered losses in excess of their theoretical book value.

First, let's assume a company with two lines of business: Line A is performing well, earning ROA of 5%, Line B is creating large book losses. In the case of E*Trade, Line A would be their brokerage and B their home equity. Or with the monolines, A would be their munis and B their ABS/CDOs.

Let's say that prior to Line B blowing up, the company's balance sheet looked like this:

Line A Assets: $80 billion
Line B Assets: $20 billion
Total Assets: $100 billion

Debt (avg cost = 6%): $60 billion
Other Liabilities (e.g., unearned premium, loss reserve, deposits): $30 billion
Total Liabilities: $90 billion

Equity: $10 billion

Now let's say that Line B loses 55% of its asset value, so it falls to $9 billion (loss of $11 billion). For the sake of argument, assume that the loss of 55% is known and no further losses are coming. If we hold Line A's assets and overall liabilities constant, we get -$1 billion in net equity.

But Line A is still performing, earning a ROA of 5%, or about $4 billion/year. Let's assume that the remaining Line B assets also have 5% ROA, or $450,000, but that the $11 billion loss is dead money. The debt is costing the firm $3.6 billion. Let's make life easy and assume the deposits or other liabilities don't have a cash cost. So the firm is still earning positive cash flow of $850 million. That cash flow has value.

But obviously this company needs some capital, particularly if they are a regulated entity that has minimum capital requirements, or an insurer who needs a certain credit rating. Let's say they need $5 billion in net equity in order to satisfy whatever requirement.

So could someone come along and buy the whole company for $6 billion? With $900 million/year in positive cash flow, the ROE on a $5 billion investment would be about 14%. If the buyer was someone with relatively low cost of capital, that investment might work just fine.

Of course, mergers and/or strategic investments aren't always about ROE alone. Some may have strategic value to a larger firm. Obvious examples would be Countrywide or Washington Mutual. When Bank of America put $2 billion into Countrywide earlier this year, it was widely viewed as a first step toward a possible merger in the future. It seems as though Bank of America viewed the convertible preferred as a cheap way of acquiring some of Countrywide's equity. This would have made a future full merger cheaper. Of course, CFC's stock has fallen precipitously since then, but that's another story.

Equity infusions are sometimes about supporting a previous investment. CIFG and Rescap are good recent examples. With financial companies, sometimes all they really need is some cash to keep the ship afloat.

Anyway, this is obviously a highly stylized example, and I'm sure you all will pick it apart in the comments, so have at it. Let me leave you with some food for thought (or commentary):

You are never bankrupt as long as investors are willing to keep funding you. In other words, running numbers and coming up with a negative net worth doesn't necessarily equal insolvency. If so, the U.S. Treasury would have been bankrupt a long time ago.

There are more forces working to keep a company going than working to drive it under. A whole host of people, from management to investors to investment bankers will work on finding solutions. With rare exceptions, no one actually working on driving a company under.

Disclosure: No positions in any company mentioned except Freddie Mac (debt).

29 comments:

Anonymous said...

AI: For the sake of argument, assume that the loss of 55% is known and no further losses are coming.

The reason some people have said buying into Countrywide or Citi is a bad idea has nothing to do with their accounting net-worth ... its that your assumption (above) is not correct.

This assumption also explains why we disagree about whether this is a liquidity crisis or an insolvency crisis.

IF the numbers were all nice and neat as in your example, then many of these companies could indeed have some value.

But there is no transparency. You don't know their net worth is ($1 billion)... If we are completely honest, the CEO of Countrywide (or Citi or whatever other example) has no clue what his/her company is worth.

What makes you think the loss is 55%? Or 20%? Or 80%? You are just guessing. I think you are fooling yourself to think its an educated guess.

First, the original information on many loans has proven to be creative writing. Stated incomes and overly optimistic (and perhaps fraudulent) appraisals mean the initial risk assessment is useless.

The basic risk model used by Wall Street assumes "independent events" -- the probability that you default is totally independent of whether your neighbor defaults. This assumption underlies VaR directly, and implicitly it underlies all the subordination/etc that underlies CDO and CMO structures.

The events are not independent. Not even close.

First, the incentives given to mortgage brokers encouraged a lot of the creative writing-- I am not accusing them of fraud; just over optimism. That means many loans are correlated by bad cashflow and collateral value assumptions. Not independent.

Second, most of the 2006-2007 vintage loans were made to persons who relied on 100% LTV, Alt-A, subprime and various other "stretches". These borrowers are all "correlated" by their low FICO scores ... they would not have qualified for a mortgage during "normal" times.

Everyone who bought in the last 2-3 years (even the high FICO scores) was buying when average home prices were at historically unprecedented multiples of average income. This doesn't guarantee default, but it does tilt the odds against the home buyers.

I could go on about more correlations, but those are the biggies.

So the whole risk minimization (diversification) that underlies all the "actuarial" models used by the mortgage industry are of little use in predicting defaults in the current environment.

The fact is, no one can really guess what the eventual defaults are going to be. And even if you could, all the securitization (and securitization squared) means you don't know who has the risk. You can't even make an assumption "on average", because the distribution is definitely not normal.

Its very rational for investors to refuse to buy paper, even allegedly "good" paper, when they don't know what the cashflows will be.

If there was a liquidity crisis, investors would not have the money to buy Treasuries (which they are doing in mob-like fashion). Investors have plenty of money, arguably too much -- but they are putting into assets where they can reasonably predict a return.

Mortgages, ALL mortgages, are unpredictable at the moment. Clearly, there is good paper mixed in there, but it is nearly impossible to sort it from the bad.

Even if you could identify the good stuff, how would you save your job? How would you convince your investors/boss that you aren't buying a basket of lemons? Heads, you give up your information (and your competitive advantage); tails you don't convince them and you are the next casualty when they pull their money.

Anonymous said...

What makes you think the loss is 55%? Or 20%? Or 80%? You are just guessing. I think you are fooling yourself to think its an educated guess.

That's what the Street is paid to do: make educated guesses and sell them. By and large it works.

Sometimes it doesn't. As a market professional, it is now generally assumed I'm an idiot. Last year, strangers at parties assumed I was a genius. Funny, I don't feel much different.

The basic risk model used by Wall Street assumes "independent events" -- the probability that you default is totally independent of whether your neighbor defaults. This assumption underlies VaR directly, and implicitly it underlies all the subordination/etc that underlies CDO and CMO structures.

Incorrect. Look up "copula" on Default Risk, for instance.

knzn said...

One way to look at it is that a business with negative net worth is a business that is operating with more than infinite leverage. If the basic business is profitable, the leverage only makes it more profitable. But leverage is risky, infinite leverage is extremely risky, and more-than-infinite leverage is even more risky, so you want to de-lever the business by investing more equity capital in it.

Anonymous said...

A good example of a company that hasn't made any money per year (per quarter doesn't count) since IPO'ing in 1998-9 and has bunch of debt and little equity is Restoration Hardware (RSTO I think). And Sears is looking at buying them, but that's another story.

Anonymous said...

With respect to the example, the "trick" or "flaw" is the assumption that the other liabilities have no cash cost. If the other liability of 30 has a short duration, simply rerun the example with assets of the A business reduced by 30 (liability paid off). Alternatively, if the other liability of 30 has a long duration and no cash cost, then its fair value is less than 30 resulting in positive equity. Alternatively, if the other liability of 30 is already discounted at a market interest rate and has a long duration, you need to set aside an increasing amount of cash flow each year to fund the other liability at maturity assuming you aren't going to default on it.

Anonymous said...

Most people think etrade sold their ABS portfolio to Citadel for 27 cents on the dollar, but in fact, it is a lot less than 27. The 20% equity stake was built
into the price.

Accrued Interest said...

I think you guys are missing my point. Gramps: all your points are completely valid. I was merely trying to show that **if** one believes they can value the troubled assets in a company, then its possible to make a reasonable investment. We can have fun debating whether Citadel is going to properly value E*trades assets or not. I believe these assets are extremely hard to value for all the reasons Gramps mentions. But that being said, 27c on the dollar is pretty low, and as other commenters point out, its kind of part of the price of getting the equity. So the 27c figure isn't really the whole story.

Right now its very hard to value many low liquidity securities, even ones where the credit quality is good. That's because there is little trading going on, and dealers are strapped for liquidity.

But that doesn't stop me from putting bids on bonds.

So I'm happy to have readers debate the merits of Citadel's move (or BoA's investment in CFC for that matter) here. I just want it to be clear that I'm not defending Citadel. I don't really have enough information on E*Trade to say one way or another. I was merely trying to show how negative book value doesn't really mean worthless.

Accrued Interest said...

Chill:

I was actually imagining the "other" liability as deposits in non IB accounts. I don't think there would be an alteration of valuation in that case?

I'm not 100% sure how an insurer's unearned premium might get revalued. I suppose that if its AMBAC, who has collected cash up front for muni insurance but counts part of that as a liability as long as the bonds are outstanding. Each year some of the unearned premium is eliminated, reducing liabilities. If AMBAC merely slowed the pace of new business, that would indeed decrease liabilities in the manner you describe.

Unfortunately, my accounting skills are woefully inadequate.

Accrued Interest said...

Jim: Exactly.

KNZN: Exactly.

Anonymous said...

In an interview, E-trade's acting CEO claimed that they had a lot of deals on the table and the Citadel/BR deal was the best. I wonder how many of us would agree if we knew the details of the other offers.

As someone who uses and likes E-trades products, I'm glad they remain independent. As a trader in ETFC stock, I took my profit and moved on. Wearing an analyst hat, I found this deal depressing. Not because it was necessarily so bad, but because it showed so clearly that people running a financial company could be smart enough to create and execute well on innovative products, but still lack even a rudimentary understanding of Risk Control 101 wrt leveraged positions.

Unsympathetic said...

1)E-trade's ABS paper was purchased at a 73% discount.

2) On Fannie's call yesterday, a price quoted to Fannie was the price Fannie used to mark their LIA loans, resulting in the required extra capital.. but the I-bank the analyst worked for wasn't marking their institutional RMBS to the same mark!

Aside from the obvious (Wall Street has become a banana republic) - how do these two actions not instantly turf everyone's positions in those paper classes?

Anonymous said...

I found this deal depressing. Not because it was necessarily so bad, but because it showed so clearly that people running a financial company could be smart enough to create and execute well on innovative products, but still lack even a rudimentary understanding of Risk Control 101 wrt leveraged positions.

I'm mystified as to what connection there might be between "running a financial company" and "Risk Control 101 wrt leveraged positions".

That they both have something to do with money? That seems a little thin.

An electronic brokerage house such as E*Trade is simply a set of computer programmes, set up to provide high security, triple redundancy (at least!), audit trails and report generation ... there's very little connection there with securities analysis that I can see.

Expecting the management of such a company to deliver superior asset management results is a bit like expecting your accountant to handle your money market investments; or expecting your stockbroker to handle both your stocks and bonds.

Accrued Interest said...

Gavin: To say that any one RMBS security is like any other RMBS security in this market is tough. So if you are saying that everyone should write down all their RMBS to 27c just because that's what Citadel paid... that doesn't work.

Unfortunately, investors are a bit at the mercy of institutions when it comes to information they disclosure and the calculations they make. Right now the problem is in mortgages, but we've had problems in the past with all sorts of asset valuations.

I think the problem is especially hard now, because some firms are going to be flat out dishonest about their positions, knowingly overvaluing things. Others are going to try to be honest, but given how difficult it is to value these securities, they might still be way off.

Accrued Interest said...

James:

I agree. E*Trade sounds like its run by opportunists. They saw the day trading thing back in the 90's. They saw the mortgage lending thing in 2006. They jumped into both, but it sounds like they didn't understand the second business at all.

Unknown said...

James I. Hymas,

Etrade runs both a bank and a brokerage. Their strategy was/is to find synergy between the two, but it was generally understood that the brokerage was raisan d'etre for the bank, and that the brokerage customers would interface with both bank and brokerage through the web 99% of the time. In many ways the bank provided fine support for the brokerage - e.g. best in class electronic bill pay service, universal (100% reimbursed) ATM access, low convenient loan offers for brokerage customers, free web-initiated cash transfers to and from other banks and brokerages, etc. All of that was a good, money making setup. The web interface to the brokerage and the trade execution/accounting platform is also considered among the best by customers. Another feature of the bank was that it offered very attractive stand alone money market rates, and Etrade advertised these rates as a draw for new customers. At some point Etrade mgmt. decided that using high leverage and reaching for yield spread on the customer deposits in both bank and brokerage was going to be a key driver of corporate earnings. We all know how that turned out. And it's reasonably clear, to me at least, that if they had even average portfolio risk controls in place, they would have averted catastrophe sometime way before they found themselves facing possible shutdown or insolvency; that is, if they had possessed a level of competence in portfolio mgmt. which was average - still not what would be required to make a good long term business from the type of yield spread chasing which they attempted - they would have come through the credit turmoil in much better shape.

Unknown said...

Sorry if it turns out I made a double post above.

I wanted to throw out another example on what I think is the theme here. The news this morning, coupled with other elements, led me to take a flier on Puerto Rico based lender, WHI. I don't have enough info to say that their survival probability is over 50%, but I think the expected value of the shares is substantially greater than the $1.20 they were trading at this morning.

Anonymous said...

AI,

Why is it difficult to value these assets? I get the no liquidity thing and not being able to get a comp, but I am not sure I understand why sophisticated investors cant value the assets. Is there no valuation model that can give you an approximation?

Accrued Interest said...

Its because every small incremental change in default rates has a large impact on cash flow. For example, there might be a security that can pay full P&I at 12% defaults, but only 50% interest and zero principal at 15%.

Throw into that recovery, which is highly uncertain, and timing of defaults, which is also highly uncertain. Both of which could make it so that a pool can withstand more defaults than would otherwise be assumed. I'm not even going to get to loan mods.

So to predict how any given pool is going to default is tough, and since small differences in losses make huge differences in valuation... you can see the problem.

Anonymous said...

So with that level of complexity its appears that investors grossly overpaid for those assets? Would that be fair to say?

Unknown said...

"So with that level of complexity its appears that investors grossly overpaid for those assets? Would that be fair to say?"

More than fair. Investors messed up a whole bunch of different things at the same time. In no particular order,

1) they underestimated the correlations in RMBS related securities due to national trends in home prices and credit availability;

2) they didn't account for the simulataneous coincidental effects of looser lending (including no/low doc loans), greater use of adjustable rates, greater use of second liens, and lagging effects of higher interest and property taxes;

3) they didn't pay attention to the difference between ratings of default probability and ratings of value when estimated recovery percentages are very different; in particular, investment grade CDOs that are less than the most senior class for a given pool often have a recovery percentage of near zero if they default -this cuts both ways, as arguably the market has recently undervalued the most senior tranches by referencing their relative value to the market prices of the lower grades, based on same state of confusion;

4) if the investors are publicly traded corporations, they didn't pay attention to what the effect of mark to market accounting rules would be on their perceptions of solvency and capital ratios if they held a lot of out of favor assets; they didn't account for the market momentum of other investors forced dumping of same assets to meet same capital reqs; almost any investment will be terrible in a period where it has a lot of weak holders.

The catch phrase "perfect storm" gets over used, but obviously a lot of different stuff is going on.

venkat said...

Hi AI,
I know you have a more interesting post already which is eliciting more comments than your earlier post. :) However, I would really like to get your response on this. I don't understand how CITI could do a deal with ADIA at such exorbitant terms. I have outlined my logic below. Please feel free to correct me. I have deliberately ignored margin costs etc to keep the argument simple

When the deal happenned, Citi was trading around $31. ADIA has effectively bought 201m calls for 37.24 and sold 235m puts for 31.83 i.e. at the money Puts and out of money Calls for the same time period. Let us assume that ADIA funds CITI by short selling CITI's stock. It can shortsell 235 m Citi's stock today which provides around 7.5 b and use those proceeds to fund Citi's debt. ADIA does not have a problem if the share price goes below 31.83 after four years. It uses the mandatory stock that it receives from CITI and covers its position completely while enjoying 11% yield. Its only risk is that the share price can go above 37.24. What can ADIA do to hedge this risk? ADIA will have a neutral position by buying 34 m calls for 37.24. How much does buying these 34m calls cost? Buying the shares itself would cost around 1b. So I would assume the options would cost even less. This is ADIA's net Investment in this deal and it enjoys 11% yield on 7bn in return for four years. Who would not invest in this deal?

Anonymous said...

AI, like you, I own a bunch of municipal bonds. One of which is insured by ACA. In your opinion, is ACA going under and should I now just assume that my bond has no insurance? I don't even know what the underlying rating is. Is that rating really hard to determine? Thanks

Anonymous said...

ACA is toast

Anonymous said...

Hymas: Wall Street genius???

Is that like the Nobel Prize winning genius that took down LTCM? Academic hubris does not equal genius, and never did.

If you visit Las Vegas or a local poker hall, most people who were not qualified to get an MBA, PhD or CFA at least have the common sense to protect the house.

Wall Street didn't. Have you noticed all your fellow geniuses taking write downs of $5-7 billion *EACH*? (some in on lump, some have a $2-3 billion installment plan). GOldman, Merrill, Bear Stearns, UBS, etc. If they are so smart, and have such mastery of risk, then why are Citi, e-trade and Countrywide begging hat in hand for new capital?

Wall Street's risk models are crap. Your comment about copula doesn't make sense, and the website you cite doesn't even list it (except in the title on a book they had an ad for).

The liquidity crisis we are facing will never be solved by dropping money from helicopters. Wall Street screwed up, and screwed up badly. Forget about the homeowners who are losing their homes....

Wall Street couldn't even protect itself.

Anonymous said...

Gramps -

Wall Street profits will be a mere $28-billion this year after all the dust has settled.

That's THIS year, the year the piper gets paid. They don't have to give back any of the profits (and bonuses!) they made riding the boom for the past few years.

GOldman, Merrill, Bear Stearns, UBS, etc. If they are so smart, and have such mastery of risk, then why are Citi, e-trade and Countrywide begging hat in hand for new capital?

It is not clear to me whether you wish to speak about Goldman, Merrill, Bear Stearns, UBS, etc., or whether you want to discuss Citi, e-trade & Countrywide. Can you clarify the matter for me?

Your comment about copula doesn't make sense, and the website you cite doesn't even list it (except in the title on a book they had an ad for).

To learn more about copulas in the context of default correlation, you can type "copula" in the search box on Default Risk - assuming, of course, that you don't want to buy the book.

Forget about the homeowners who are losing their homes....

You know, it seems everybody is talking about poor but honest Dickensian homeowners who were tricked into ruinous mortgages that are now going to ruin the financial system.

I'm not so sure about that. There is some evidence that the archetypal subprime borrower is actually a rather cynical speculator who was utilizing the non-recourse feature of American mortgages to, essentially, take a call option on a house. You could also look at it as 100% non-recourse margin on an illiquid security, if you wanted to.

Do you have any figures on just who is getting foreclosed?

We used to have non-recourse mortgages in Canada, but the banks wised up. *sigh*

Accrued Interest said...

Citi and Abu Dubai: I think Citi did the math. They need a GIANT cash infusion. If they went to the market asking for $7.5 billion, it'd be extremely expensive. Freddie Mac paid 8 3/8 for $6 billion, so Citi would probably have to pay 12!

So right or wrong, they must have thought Abu Dubai was the better deal. Maybe there is also a side agreement to do more if Citi needs it?

On Wall Street: The basic model is to ride the waves while its good, then take your lumps when it falls apart. I agree with James in that the model works. Gramps, you might have an ethical objection, or perhaps argue that WS is making things more volatile for every one for their own benefit. I am sympathetic with those points, but I'm not sure there is a better system.

ACA: In a LOT of trouble. I'll tell you that ACA bonds are trading like there is no insurance. Maybe worse than that.

Anonymous said...

The part I like best about the Wall Street model is that it encourages entrepreneurial activity.

I have never worked for a Wall Street firm, so my information may have come through rose-coloured filters, but it is my understanding that, if you are professional with an idea, there is a very well defined process to making it work. There is ready access to capital; if the idea works, the company makes some money and the professional gets a fat bonus. If it doesn't work, the company loses some money and the professional gets fired.

Do you believe in your idea? How much are you willing to bet?

In Canada it's more a matter of proceeding through channels (comprised of people who won't see any rewards for their approval of a good idea), reaching consensus and diluting the rewards to the originator considerably. The banks, having taken over the industry about 20 years ago, are really starting to impose their culture even to the level of front-line salesmen.

Anonymous said...

I don't believe "Wall Street" really made $28 billion this year.

If you dig into the numbers, they made a ton of money in M&A -- advising on mergers and take-overs (or defense thereof).

Fixed income underwriting is deep in the red almost everywhere -- both from write downs of warehoused loans and from 6 months of basically no activity (but much of the fixed costs including payroll).

Trading operations are deep in the red everywhere except Goldman, and one desk at Deutche.

Hymas: if you had spent a few minutes during your career on a sell side trading desk, you would know that most of them have trading strategies that are essentially short volatility (much like selling a strangle). If there is big volatility, as there was this year-- they have to scramble to hedge themselves and best they can hope for is to not lose too much. On a mortgage desk, there is an explicit straddle being sold (the mtge is long a bond and short calls, then they hedge by selling a bond).

Govt bond desks probably did ok, as they generally combine a short vol dealership function with a gamma (convexity works in their favor for big moves)... but given the size of markets, its govt bond desks could not hope to make money as fast as the credit desks (mtge and corporates) were loosing it.

Anonymous said...

Hey AI,
Have you seen this SEC filling about Etrades portfolio? http://www.sec.gov/Archives/edgar/data/1015780/000115752307009928/a5520447ex99_2.htm

While most of the loans look good acording to FICOs, largest number (and rightfully so, since most of the loans are in the top 2 tiers) of deliquent loans is coming from the top tier of FICO scores.

Does this give you any more insite into etrade and what might happen to them?