(02/07/2004)
SourceMore on 990N:
I was the person who went to the floor of the Mercantile Exchange on a fact finding mission. Having studied the S&P price action in every conceivable fashion, I wanted to find out why movement on the S&P had changed so much over the last 18 months. I (and others at a large hedge fund) first noticed a significant change in the trading patterns around July of 2002. After a month, the patterns returned to normal, as the market promptly went back down. Then around September 23rd (a fed meeting), we immediately noticed the trading anomalies returned, as the market promptly went back up. This is what we observed:
1. Dow 50 declines in a hour
Irrespective of the selling pressure, the Dow would not decline much more than 50 points during any one trading hour. Once it was down that much, volume of the futures would plummet (as if program trading restrictions had been triggered). (BTW; if you look in Barron's, you will notice that all domestic investment banks have now shut their program trading operations down. I wonder why they would shut down a huge profit center? Now, the largest index arbitrage operation is now ABN Ambro.) Then, after an hour, volume & price movement would pick up again as if the collar had been lifted. So, initially we thought that the Fed & NYSE had altered the program trading collars.
2. Overnight futures
In previous years, the overnight futures action was almost always wrong. Or, at the very least, the overnight price action would return to even (close the gap). Now we have so many unfilled gaps that I have lost count. However, current trading is more like someone is trying to obtain a price goal with the least numbers of futures contracts. So, bid the price up overnight, and maintain the price by sitting on the bid. There is no intraday ebb and flow whatsoever.
3. Managing the price
Once the price has been thrown up, the price level is maintained by sitting on the bid. At other times, the price is collared by putting a huge number of bids and asks on the screen. This discourages one from getting long or short. Often these orders are pulled unfilled. However, they are an effective deterrent from entering the market one way or another. Since when is most of the price on the S&P determined overnight? Any decline that is allowed to happen from these bid levels all but impossible to trade. If you short, the "inevitable" surge back to those levels is lightening quick as all the shorts cover.
4. Jam jobs
All of this leads to the number one anomaly. The massive surges in volume we have labeled "jam jobs." Due to the lack of volatility and predictability in the S&P, the futures volume has slowed to a trickle. Thus, it has become easier and easier to maintain the price (and makes the manipulation all the more transparent). No one will short anymore for fear of the "jam jobs." This occurs when, out of the blue, when trading is slow, someone drops market order for thousands and thousands of eminis. (we call it a "volume bomb." It's a testament to how raw, and obvious, the manipulation is that we have a label for the routine. If you were actually trading the S&P, you would too. In fact, we find it very much like the movie Groundhog Day, every day the same tricks are used.) In the span of under 5 minutes, the price action will move (up, never down) several points.
The net effect of all this, is that shorting is impossible. Perhaps all of this does sound nutty, but if you are trading the S&P, you are dealing with this every single day. Rather than lazily dismissing something offhand, I encourage anyone to contact somebody who trades the S&P eminis (preferably on Globex). Then, ask them about the price action and the counterparty 990N.
5. Proof of price manipulation
The best proof of price manipulation is the price movement. We haven't had a 2% down day on the S&P in 278 trading days. Look at the change in the statistics of downside volatility:
-1.00% -2.00% -3.00% Price % Decline Trading
Days
2004 121 11 - -
2003 250 58 13 3
2002 251 86 33 10
2001 248 67 16 4
2000 252 73 25 7
1999 252 67 15 1
1998 252 62 15 4
1997 252 61 11 2
2004 100.00% 9.1% 0.0% 0.0% % of Days
2003 100.00% 23.2% 5.2% 1.2%
2002 100.00% 34.3% 13.1% 4.0%
2001 100.00% 27.0% 6.5% 1.6%
2000 100.00% 29.0% 9.9% 2.8%
1999 100.00% 26.6% 6.0% 0.4%
1998 100.00% 24.6% 6.0% 1.6%
1997 100.00% 24.2% 4.4% 0.8%
Moreover, over the past 18 months we have had a slew of historic upside movements. Twice in the last 18 months, we have had 8 straight up days in the futures. (And up 7/8 days once.) This has not occurred once in the last eight years (other than the two mentioned). During one run, we were up 24 of 27 trading days, which occurs about once a decade since 1940. In May, the futures were up 12 of 14 days, which occurs about once every 5 years.