Skip to main content

Timberwolf Continues To Stalk Goldman Sachs

  • Author:
  • Updated:

Financial product salespeople, if they know what's good for them, should be thankful for car dealers. Not used car dealers, either, new car dealers: because of the world's familiarity with their business model, if you sell a client a product at 100 and then tell them the next day that it's worth 95, you have at least some outside shot at pacifying them by explaining, slowly and patronizingly, "it's like buying a car: the price drops as soon as you drive it off the lot."

I mean, that's true of buying a toaster or a bunch of carrots, too, but nobody marks those to market, so. I guess people do mark their cars to market? That seems to be a thing. In any case, "mumble mumble mumble drive it off the lot" sounds much better than the alternative, which goes something like "yeah, we thought it was worth 95 but we sold it to you at 100, problem?"

Remember Timberwolf? Timberwolf was an RMBS CDO that Goldman Sachs marketed. It was also "one shitty deal," in Tom Montag's immortal words, and some of it was sold it to some Australians with the buzzword-salad name Basis Yield Alpha Fund (Master), and Tom Montag was right, so, that worked out poorly for Basis Yield Alpha Fund (Master). Working out poorly was a feature of a lot of Basis Yield Alpha Fund (Master) investments; before they bought Timberwolf they bought another MBS CDO called Point Pleasant, also from Goldman, and whereas a Timberwolf will of course rip your face off - that's just evolution - the face-ripping they experienced from a Point Pleasant seems to have come as some surprise.1 Anyway, they sued, and while Goldman has engaged in marvelous jurisdictional kerfufflery that got it tossed from federal court, they are still in New York state court, which refused to toss it late last week.

Here, from the opinion refusing to toss the case, is Basis Yield Etc.'s core allegation:

BYAFM's core allegation is that Goldman knew that Point Pleasant would decline in value and that Goldman's strategy was to offload Point Pleasant onto its clients to avoid taking a loss.

This is a boring core allegation! Also it is impossible! Point Pleasant was a bag of mortgages. Goldman does not employ psychics. It cannot be that "Goldman knew that Point Pleasant would decline in value." Goldman may well have thought that, it may well have positioned itself to profit from the decline in mortgage prices, and it may have sold an MBS CDO as part of that positioning, but it couldn't know that Point Pleasant would decline in value.

It could on the other hand know another, not unrelated thing. Go ahead and feel the panicky awkwardness in this paragraph:

On April 17, 2007, BYAFM purchased Point Pleasant securities with a face value of $15 million (with a unit price of $81.72) for $12,258,000. BYAFM and Goldman entered into a Repurchase Facility under which Goldman could make a margin call if it decreased its mark (unit price) on the Point Pleasant securities. On April 27, 2007, Goldman made a margin call on BYAFM for $3.2 million, $3.091 million of which was attributed to the Point Pleasant securities. When BYAFM questioned the margin call, Goldman revised it down to $700,000, with approximately $600,000 attributable to Point Pleasant, reflecting a drop in the unit price of the Point Pleasant securities to $76.72. After further discussions between Goldman and BYAFM, Goldman further revised the margin call to $35,000 based on the representation that the decrease in the unit price was due to a system error. Shortly thereafter, Goldman reported that there was no margin call and that the unit price was still at the purchase price level. In response to an email on May 21, 2007, in which BYAFM inquired about Goldman's valuation of Point Pleasant, [GS salesman] Maltezos wrote "For end April, these were marked at the same level [BYAFM] bought the bonds at, i.e. 82-23 (81.71875%)." On June 12, 2007, Goldman made another margin call on BYAFM and indicated that the unit price had decreased to $75. Maltezos told BYAFM that "the 75 mark for end-May is the first adjustment we've made since you bought the bonds." BYAFM alleges that Goldman did previously lower its internal mark on the Point Pleasant securities, but withdrew the margin calls to give BYAFM a false sense of security so that Goldman could induce BYAFM to buy more toxic securities that Goldman sought to offload."

That sounds terrible! Perhaps I am naïve? But the cynicism of BYAFM's cynical allegation there - GS lowered its mark, then raised it in order to make BYAFM feel better and buy more CDOs (including Timberwolf) - is less cynical than my cynical assumption, which goes like:

  • GS had thing marked at somewhere between ~62 and ~77,2
  • GS found buyer of thing at ~82,
  • GS sold thing at ~82,
  • on margin;
  • GS kept thing marked at ~62 to ~77,
  • GS issued immediate margin call for the 20 point difference, which it then walked back to a 5 point difference.

With the ten-day delay being cosmetic or systems-related. And then, of course, the awkward phone call between the guy in collateral ops and the salesman who is not happy that a margin call just alerted his client to how much they're being ripped off:

Salesman: We have to reverse this margin call.
Collateral guy: I see it at 77, and that was moved up from 62 a few days ago. We're calling based on 77.
Salesman: My clients will literally murder me for overcharging them.
Collateral guy: I see it at 77. We're calling based on 77.
Salesman: Also they will never do a trade with me again.
Collateral guy: I see it at 77. We're calling based on 77.
Salesman: Also we're axed to offload an even more horribly named CDO on them.
Collateral guy: I see it at 77. We're calling based on 77.
Salesman: I mean, it's named after a wolf. A wolf!
Collateral guy: I see it at 77. We're calling based on 77.
Salesman: Do you have a supervisor I can speak to?

This guy Greg Smith wrote a book about Goldman, you may have heard. One thing he complains about in his book is a junior salesman who gloats about selling a product to a muppet client who doesn't price-check him and so overpays for the product by $2 million. It is important to realize that this is also impossible in the strict sense, just because you can't "overpay" for a non-trivial financial product "by $2 million." You can pay $2 million more than someone else might have charged, or you can pay $2 million more than Goldman's mark, but you can't pay $2 million more than the fair price because there's no fair price. The mark - Goldman's best estimate of fair value - isn't the fair price, because if it was Goldman would never make any money, and you wouldn't want that would you?3 You can pay an unfair price - if the mark is $77, $200 seems unlikely to be fair - and you can rough out market rules of thumb for appropriate markups,4 but you can't really quantify the unfairness with any precision.

But there is a mark! That description of Goldman's margin call and subsequent walk-back is not, like, Proof Of Fraud - even if it's true. (I have no idea if it is; perhaps Goldman's explanation of the slowly vanishing margin call - just a series of unrelated technical errors - is right. If so, um, fix those computers guys!) Goldman is in the business of charging more than fair value for products, because that's how it makes a profit; if it sold everything at cost how would it pay 50% of revenue out in comp? There's nothing wrong with that, any more than there is with selling carrots or cars for more than their cost.

The margin call seems fine - maybe less fine if it was for 20 points - but the vanishing isn't great. Selling something at 82 and saying forthrightly "yeah, we took some profit, now it's worth 77" is certainly awkward. Marking it back to 82 probably avoids the awkwardness and makes the client feel better in the short term, but there's a cost. For one thing: it does look fraudy in hindsight, especially if you keep selling the client more CDOs. For another, though, it creates the impression that you're in the business of doing something other than charging clients more for securities than you think they're worth. If clients know you're charging them a markup, then they can do things like ask "what is the markup?" and then argue about it. If you just tell them that their product is worth what they paid for it, you lull them into trusting you. And if you do that, you can't act too surprised if they end up feeling betrayed.

Goldman loses bid to end lawsuit by hedge fund over CDO [Reuters]
Goldman Sachs and the sophisticated investor: Who's duping whom? [Reuters / Alison Frankel]
Basis Yield Alpha Fund v. Goldman Sachs [via Reuters]

1.I once worked on a deal code named "Project Easy Living." To no one's surprise, it wasn't. A colleague suggested we call the next one "Project Enjoy Your Weekend."

2.The opinion notes the $600k margin call reflects a drop in unit price to ~77, but of course the initial $3.1mm margin call reflects a twenty point drop in value. Which seems big for ten days! Though, I mean, April 2007, maybe? I can't get early-2007 ABX historical data on my cheapo Bloomberg but page 11 here suggests that a 20, or even 5, point ten-day drop in subprime RMBS then would have been unusual.

3.This is why it's so strange that Greg Smith thinks it's a smoking gun that banks sometimes make money trading every day for a whole quarter. "Trading" is - one hopes! - market making and charging a markup for providing liquidity: you're supposed to make constant steady money. It's like saying "holy shit, Apple must be ripping you off, because they make a profit every time they sell an iPhone." If you only make a profit selling iPhones 40 out of 60 days, you're doing it wrong.

4.Is 82 fair? If, like, a NYSE listed stock was trading at $77 and I sold 1,000 shares to you at $82, no, that's not fair, or close to fair, and there is a rule of thumb that a few pennies is a fair markup but $5 is unfair by rounds-to-$5. In some illiquid products 82 vs. a 77 mark is surely just fine.