Friday, 12 March 2010

Mashreqbank v AlGosaibi – Analysis of FX Deals

This post is based on the affidavit (including appendices) submitted by Bruce R. Grace, Esq. of Baach Robinson & Lewis PLLC, who are the attorneys for Ahmad Hamad AlGosaibi & Brothers Company ("AHAB") in the above lawsuit. Attorney Grace submitted these documents to the Supreme Court of the State of New York on 5 February 2010. They were filed as Document #83. But since then for some reason there's been a change. Document #83 has been deleted and it and each of the exhibits contained have been filed as separate documents all dated 25 February. The specific document used as the basis for this post and referred to herein is Document #87-1. 

You can access electronic documents filings for this case at the Supreme Court of New York's website (http://iapps.courts.state.ny.us/webcivil/FCASMain). Search using Case Index #601650/2009. Follow the steps. You have to go through several pages. On the page labeled "WebCivil Supreme – Case Detail" go to the bottom of the page and click on the "Show EFiled Documents" button.

As usual, a caveat. As you read documents filed in the case, keep in mind that each side's lawyers are not disinterested partisans of truth. They are hired to represent their clients.

From the documents I've read it's clear that both parties agree that there was an FX transaction in the name of AHAB with Mashreqbank under which Mashreqbank was to pay US$150 million to AHAB's account at Bank of America on 28 April 2009 and AHAB was to pay SAR564.3 million to Mashreqbank value 5 May 2009. They also agree that Mashreq fulfilled its side of the contract but that AHAB did not.

As you'd expect both parties have different views about this latter fact. Here's how I would summarize their arguments. AHAB has two key contentions. First that they were the victim of a fraud perpetrated by Mr. AlSanea. And second that from the nature and pattern of the deals Mashreq should have known the deals themselves were questionable. Therefore, Mashreq is in some sense complicit. On its part, Mashreqbank has vigorously denied knowledge of any wrongdoing in the deals and asserts that it was acting in commercial good faith. The Bank also argues that the instructions it received appeared to come from duly authorized parties at AHAB. Simply put, there is a commercial deal which AHAB needs to honor. That's my take. But, I suggest you read the case documents themselves to get a complete picture of their positions.

Putting aside the issue of who is right and who is wrong, let's take a closer look at the FX deal which is the subject of this case.

The FX transaction is somewhat unusual as it is a "split value" deal. Normally FX deals have both parties making payments on the same day. If Bank A sells Bank B Sterling against the US Dollar, on the settlement date Bank A remits Sterling to Bank B's account. And on the same day Bank B remits the countervalue in US Dollars to Bank A's account. In a split value date deal, one party pays before the other. The party who pays first is taking risk that the second party might not pay. It is taking a credit risk that the second party may not pay its side of the transaction due to financial problems – being put under administration, entering bankruptcy, etc. Even in an FX transaction where settlement occurs on the same day, there is that credit risk. Herstatt Bank is an example. However, the longer the gap between the first party's payment and the second party's the more credit risk the first party is bearing. 

So what could be the reasons why parties would agree to a split value settlement mechanism?

Credit Risk Management

If one of the parties were concerned about the creditworthiness of the other, it could mitigate its risk by requiring the weaker party to pay first. After confirming receipt of funds and only then, the stronger party would remit its funds. This eliminates credit risk on the weaker party. The time between the two payments would be primarily a function of two things. First, the time it takes for the stronger party to confirm receipt of the funds from the weaker party. Second, how soon thereafter, the stronger party is operationally able to make its payment of the countervalue currency.
Given the currencies involved – the US Dollar and the Saudi Riyal- confirmation of receipt should be relatively easily. There are two widely available methods which provide quick and efficient information (as well as payment functionality) to banks around the world. The shared global "utility" called SWIFT. And proprietary systems offered by various major banks around the world – generally based on access via the Internet using PCs.  AHAB and Mashreq would have access to one or both of these.

In terms of payments, same day payments for these currencies should be no problem.  The US system has been "wired" for some time.  Saudi Arabia has had an electronic interbank payment system - including same day payments - since 1997 SARIE (Saudi Arabian Riyal Interbank Express).  Payment instructions could be transmitted by SWIFT or a proprietary bank system.

Two conclusions. 

First, This can't be a credit motivated transaction since the stronger credit, Mashreq, is paying first.

Second, the gap between payments, seven days, is longer than what a credit driven transaction would require.  Three days is probably sufficient as this represents the  "worst case" in terms of operational timing.  That is, if  Mashreq were to remit US Dollars on a Wednesday,  because of later NY working hours, AHAB might not be able to confirm receipt until the next business day. Even if staff came in on Thursday to confirm receipt, a SAR payment couldn't be made until Saturday. 

The only thing that could lengthen the period would be seasonal holidays, e.g., Eid Al Fitr, Eid AlAdha, National Day, etc.  Were there any seasonal holidays during late April / early May 2009 that extended the time period?   No! 

So I think we can safely exclude credit concerns as a motive for a split value date deal in this case.

OPERATIONAL REQUIREMENTS – DISPARITY OF WORKING DAYS

The second reason for these transactions is ostensibly disparity of working days. That is, on the settlement date, New York is open for US Dollar payments but Riyadh (and the rest of Saudi) is closed for the weekend or a holiday. As noted above, Saudi Arabia's "weekend" is Thursday and Friday. 

In this transaction US Dollars settled on Tuesday 28 April 2009 with the SAR on Tuesday 5 May 2009. There were five Saudi working days from 28 April to 5 May: Tuesday 28 April, Wednesday 29 April, Saturday 2 May, Sunday 3 May and Monday 4 May. There were no "seasonal" holidays during this period. So it is hard to see that there is an operational reason for the seven day gap.

A key question though (at least in my mind) is whether even with a disparity of working days such a split value deal would be entertained. 

When I worked for financial firms in the ME managed according to "USA" practices, a split value deal (except one undertaken for credit reasons) was strictly forbidden.  And generally there was  no interest in the extra hassle involved with credit motivated split value deals.   Two reasons. First, such a transaction could be a way of providing a short term money market loan to a party. Such extensions of credit were required to take place under lines specifically designated for loans.  If a transaction does not appear as a loan, a bank might extend another loan well beyond its risk tolerance for that customer. Another reason was integrity of our financial statements and regulatory reporting. A loan should be reported as a loan not an FX transaction.

Now I recognize that contrary to the beliefs of some what a "USA-style" managed institution does is not necessary infallible behavior. So I checked with an old friend who is long experienced  in Treasury at the DGM/AGM level in both "Arab" and "European" style managed banks. He told me that none of the institutions he worked for ever would entertain such a transaction. The counterparty would simply be told  that the deal had to be done on a common working day for both currencies so that settlement of both sides of the transaction could take place on the same day. But perhaps our horizons are too narrow.  Maybe other institutions apply different rules.  Even if that is the case (and I'm not persuaded that routinely doing split value date deals makes good business sense), the fact is that 28 April was a valid business day in both the Kingdom and the USA. So there was no operational reason both sides could not have settled on the 28 April. 

My friend also commented that the 3.762 FX rate used in the transaction was "non reflective of the market price". In his banks and mine, it was strictly forbidden to book transactions at non market rates. The concern was that one was assisting someone in manipulating their financials.

RATIONALE FOR THE SPLIT VALUE TRANSACTIONS

So if we've eliminated these two justifications, what could be the rationale this transaction?

If we examine the pattern of transactions, we can perhaps gain an insight into their purpose and rationale. 

In his affidavit Exhibit 1, page 20 paragraph #21 Attorney Grace states that between February 2005 and 5 May 2009, Mashreq and AHAB engaged in over "100 purported 'split value foreign exchange transactions' substantially identical to the transaction pleaded by Mashreq in the Complaint. Between January 2008 and 1 May 2009 alone, 52 split value foreign exchange transactions were carried out between Mashreq and the Money Exchange, totaling US$4.7 billion." 

He then goes on to describe that the transactions involved a payment of US Dollars by Mashreq to AlGosaibi's US Dollar account at Bank of America New York with AHAB to pay SAR to Mashreq at National Commercial Bank 3 to 12 days later. The FX rates used in these transactions were not at the prevailing spot FX rate. The SAR is fixed to the US Dollar at 3.75 SAR = $1.00. Interbank trading takes place in a narrow range around that rate. Before the settlement date, AHAB and Mashreq would engage in an offsetting deal to roll the existing deal forward. He also noted that in every deal Mashreq made a profit. AHAB never did.

Here's an outline of how this scenario might work. First, there is the original deal. Let's use the deal settling 5 May for SAR as the "Original Deal" (though to be clear that deal was actually the rollover of an earlier deal). That deal involved US Dollars against SAR with Mashreq paying US$150 million value 28 April with AHAB to pay SAR564.3 million on 5 May. If we presume AHAB doesn't have any money or wants to use its money for something else, how does it settle the payment due on 5 May to Mashreq?

Let's treat this as two separate deals. First a deal to cover the SAR payment due on 5 May (a "Closeout Deal"). And then a second deal to re-open the position in US Dollars (a "Rollover Deal").

In the Closeout Deal AHAB buys SAR from Mashreq against its (AHAB's) payment of US Dollars to Mashreq with settlement of both for value 5 May at the Spot Rate of 3.75. That takes care of the obligation to pay SAR from the Original Deal, except for Mashreq's profit SAR0.9 million which AHAB has to fund from its own resources. (The "profit" arises from the difference in SAR betweens US$150 million at SAR 3.75 = US$1.00 and SAR 3.762).

But now it needs US$150 million to pay for the SAR value 5 May under the Closeout Deal. Where does it get the US Dollars, if it doesn't already have them? From the Rollover Deal. Let's assume the Rollover Deal has the same terms as the Original Deal except the value dates are different. So, AHAB would buy US Dollars from Mashreq value 5 May against a payment of SAR to Mashreq value 12 May. This deal re-introduces the split value date.

While the Affidavit it sounds as though the Closeout Deal and the Rollover Deal were combined into a single deal, the Exhibits contain a copy of AHAB's confirmation for the Original Deal (itself a Rollover of an earlier deal). That confirm shows that there must have been two separate deals as outlined above. AHAB and Mashreq could agree to net the two deals' payments. And thus AHAB would owe Mashreq SAR0.9 million.   AHAB's obligation to pay Mashreq US$150 million (Closeout Deal) would be "offset" by Mashreq's obligation to pay US $150 million to AHAB (Rollover Deal).

Putting this information in tabular form might make the explanation clearer. The table below summarizes the cash flows by value date that would have occurred from a rollover of the transaction maturing 5 May. But note: no such rollover occurred: no Closeout Deal and no Rollover Deal. And since AHAB didn't settle the SAR payment on 5 May, Mashreq is pursuing them in Court.


TransactionUS$ 28 AprilUS$ 5 MaySAR 5 MaySAR 12 May
Original Deal

FX Rate = 3.762
+$150MM -SAR564.3MM
Closeout Deal

FX Rate = 3.75
$0-$150MM
+SAR562.5MM
Rollover Deal

FX Rate = 3.762
$0+$150MM
-SAR564.3MM


 
TOTAL CASHFLOW

TO AHAB
+$150MMUS$0-SAR1.8MM-SAR564.3MM


As you can see from above the Closeout Deal and the Rollover Deal effectively push forward AHAB's payment of the SAR564.3 million to settle the original inflow of US$150 million. The only payment that does occur is the profit payment to Mashreq on 5 May. On each settlement date in the future (assuming the deal would be continued to be rolled forward), AHAB would pay Mashreq its profit. That explains why Mashreq always was making a "profit" on each deal.

Based on the above. 
  1. These FX transactions were the equivalent of short term loans. 
  2. The persistent "profit" on the deals was in effect the interest payment on the loan. 
  3. The transaction FX rate of 3.762 compared to the 3.75 fixed parity results in an implied borrowing rate of 8.32% per annum. That rate might seem high. The day before the 28 April value date one month Libor was trading at roughly 43 basis points. That means the margin over Libor was roughly 7.9%. But remember that at that time spreads were still elevated due to subprime crisis, the failure of Lehman, etc. So there is some rationale to the credit spread here. 
  4. Also as is hopefully clear, the Rollover Deals did not result in AHAB getting more money from Mashreq. Rather the Rollovers merely extended the maturity of the loan. So despite the US$4.7 billion volume of transactions mentioned, AHAB only received  (borrowed) US$150 million from Mashreq. 
  5. Mashreq's dealers must have known that they were extending loans. Whether its management did is not determinable from the documents I've seen. 
  6. In terms of justification for the transaction, it's hard to imagine that anyone with a knowledge of AHAB's business could consider these transactions were commercially necessary to fund the needs of the Money Exchange business. The volumes are too large relative to the Exchange's business.  Since the amounts were rolled forward, there was only one net cash inflow to AHAB. 
  7. Nor would these transactions likely to be financially motivated "hedges", particularly, since the US$/SAR FX rate is fixed (and has been so for many many years) and since the Saudi Government has ample financial resources to maintain the "peg".

3 comments:

Unknown said...

most interesting articule.
need to reach the writer. how can i do that?

Anonymous said...

There is an error in the calculation it should be $ 150 * 3.762 = 564.3. you have mentioned 563.4. The net profit that AHAB would have to cover would be sar 1.8 mio instead of sar 0.9 mio as mentioned. I love the article...

Abu 'Arqala said...

Anonymous

Indeed there is an error. Now corrected.

While I'm embarrassed to have left money on the table (even if it belonged to Mashreq), at least I can take comfort in the fact that people actually do read what I write.