Analysing Sumitomo
   
   How to lose $1.8 billion
   


















 

Analysing Sumitomo

How to lose $1.8 billion

The first eight items in the list are clues that a problem was likely to occur. Many items on the list are classic signs of a trader operating out of bounds and should have been seen by the people involved at the time. The last two items point in the I I I direction of those doing something they should not have been.

After uncovering the bare facts, lets look at how it's possible to lose $1.8 billion undetected, shedding light on the shortcomings at Sumitomo.

Price movement flat

In order to accomplish the alleged market corner, it would be necessary to control, that is to own, most of the free stocks without putting them back into the market on loan. If Hamanaka owned 400,000 tons or so of stocks outright, he might have lost around $900 per ton on these or about $360 million.

Financing costs

These stocks would cost about $1.1 billion to purchase outright and this would have to be borrowed directly from banks or financed through prepayment swaps. The annual cost of financing this position is at least $50 million. If there was in addition a price exposure through the swaps the number could reach $150 million. And in addition to the cost of financing the longs, Hamanaka would have needed to finance any losses being carried forward. This could have cost him $10 - 100 million per year.

Basis movement

Next is the basis. Hamanaka would have had a significant exposure to the cash futures basis if the estimates of his forward sales hedged in futures and options are valid. If these positions have been properly marked, the loss could have been around $200 million.

Gamma Risk

Last and most significant is gamma risk. If in fact Hamanaka hedged large portions of his forward sales by selling out of the money puts, these positions would have exploded when copper prices plunged in the spring. Based on the size of the reported futures positions unwound, his reported forward sales, an educated guess of the change in his delta hedge, and the price drop, these losses could have reached $700 miIlion in May and June alone. This is particularly true if after he was reassigned, no one was watching it.

This totals at least $1.110 billion since January. The total is not an effort to state the positions exactly. It is only to show that it is entirely plausible that the $1.8 billion loss occurred largely this reporting year. One does not have to wonder how you hide $1.8 billion for ten years.

Structure without detection

So how did Hamanaka hide losses? Sumitomo has said he used unauthorized accounts at brokerage firms. We should not however assume he hid everything. It is possible that portions of the risks were on Sumitomo's books. Sumitomo has called these the authorized deals. Authorized deals could have lost several hundred million dollars.

The unauthorized losses are always referred to as transactions that abused the name of Sumitomo. This certainly infers that the brokerage firms with power of attorney were doing highly leveraged transactions in the name of Sumitomo, while Sumitomo thought they were trading cash and futures with the brokerage company. It is plausible that senior management didn't realize they were financially backing some trades especially if they involved high leverage and mathematical formulas. The senior management could easily have been incapable of understanding the magnitude of the risks. Management could have signed documents without reading them.

But hiding cash payments is more difficult. Sumitomo has said that Hamanaka falsified some reports. Perhaps he falsified cash reports. In fact we know that it was ultimately a cash payment that caught him.

It is quite possible for me to see how all of this could happen undetected. And further, it is possible to see how the statements being made may not really conflict. If the statements are read carefully, they seem to refer to different actions, different parts of a huge problem.

Let's now move on to a list of seven problems that are important as well as obvious.

Continue

Case Studies * Analysing Sumitomo