[WARNING: Idle Speculation Ahead]
Last week, I mentioned what I suspected was potential spoofing in the ED markets. Of course this was just a feeling. I don’t have the tools to record and monitor the bid-offer quantity data. However, I did run across this article in ZeroHedge, on how a trader was fined for trying to manipulate the algos. The gist was that someone put in large buy orders, trying to get trend-following algos to come along (and maybe even raise the bid). He would then proceed to cancel his orders and sell to the algos. He was punished, but the algos were not.
The street-justice side of me thinks… if some algo is going to try and front-run large orders, then from a game-theory perspective, you need to be able to smack them down enough to create a more level playing field. Otherwise, it will be difficult to get filled on your larger orders at your desired price. It’s stupid to allow algos to jump in front with no penalty. If the move is correct, they could make a few pips (especially if the larger order chases and raises their bid), and if they are wrong, they know they can sell back out to the traders they front-ran, presumably at just a pip lower. It’s like asking some people to play poker with their cards face-up. The eye-for-an-eye way to achieve game theory balance is to allow people to put in large buy orders when they want to buy OR sell.
The more logical side of me thinks… this brings up the problem of fake orders and spoofing. You run into the issue of market participants putting in all kinds of orders, when in fact they are doing so just to manipulate the markets (or algos). The guiding ethical trading rule seems to be that market participants can only put in orders they intend to get filled on. Front-running still seems sketchy, but I suppose those algos do intend to purchase those orders. Weighing all of the above, the best legal thing to do is to be a little more deceptive about the sizes you want to work (so as not to trigger the front-running algos).
But let’s get back to trading in practice. There are many shades of grey in terms of how orders are worked. Many seemingly “unethical” practices are common in futures trading:
- The practice of bidding more than you want, so you can get a larger allocation. This practice is so common most traders may not give it a second thought. But if you think about it, it’s a shady practice. You see someone bidding on 10K FFF8, and since this doesn’t move much, you put in a bid for 100K to get a larger allocation of any fills, even though you may only be looking to buy 500. This takes away difficult-to-get fills from the genuine bidders. Some may call this “stealing” or “line-jumping.” But it’s a similar shade of grey as front-running. I suppose the CME loves this, since it adds depth, so they may be more willing to look the other way. I don’t think you could argue that a trader wants the entire 100K – you can see this as the bid size decreases quickly as small lots trade bid-side. But you do it to game the system. The next level is when you know people are doing this, so you feel like if you don’t do this, you put yourself at a disadvantage to the markets in a zero sum game, and before you know it, everyone is doing it. This sounds like it should be some kind of violation (since you don’t want it all), but it happens all the time.
- As the markets start leaning away (say sells off), the offer size grows. Any sale on the offer side becomes that much more off-market, so the algos want a larger piece of this. Again, you write these routines to game the system. You do this to get a larger allocation of stray trades that may be fractionally off-market. You may or may not want the full size, but I would guess that most times, the algo would not.
- A substantial percentage of the bid or offer size comes from “implied” orders. There are many times where people may want to buy calendar spreads, butterflies, options hedges, etc, whose price would feed into the individual contracts. Lately though, I seem to be finding large implied bids and offers that can not be traced back to “common” structures. Maybe I wasn’t looking carefully enough. But it wouldn’t surprise me if some/most of the implied orders were being manipulated to show larger offer sizes to nudge the markets a certain way.
- When something looks less likely to trade in quiet markets or the settle, you spoof the markets. I suppose there is no reason to beat around the bush and use implied offers (which may be more insulated). If you have a large enough position and you want a good mtm for the day, you just offer large size, and the trend algos follow. Showing large size may even help you unwind your position via good flips, keeping your overall position size stable. If you do this often enough and some time-algo catches on, they just hop in and start offering for you at the designated time. How is anyone going to prove you are trying to manipulate the markets? The fact that there have been so few prosecutions speaks to how difficult intent is to prove.
Ever since the pre-algo stone age of trading, we had large players who would constantly keep the markets leaning a certain way (or wanting closes a certain way). This may have been a longer term view, or they wanted to squeeze the markets, get a good daily P&L, or whatever. This is as old as trading. A few years ago, we had Bill Gross with seven-figure ED contract positions constantly keeping certain parts of the curve bid. Some of the larger locals would regularly mismark some of the back EDs. To the above, if you add things like unchecked trend-following algos and potential spoofing algos, then you are adding gasoline to the fire.
This reminds me of the steroid era in baseball – formerly large players being aided by trend algos. Everybody knows the rules, but when money is involved, people gravitate towards ways to bend/break the rules. What happens is, the rule benders profit at the expense of the rule followers (since futures trading is a zero-sum game). Some breakers will get caught and punished, and some benders will get prosecuted. However many will not. The authorities should constantly be looking to ensure a level playing field.
But this could take forever (or never). So in the meantime, the rest of us should be more vigilant and cognizant of these things going on in the markets. It becomes even more important to identify what you think is going on in different parts of the curve. You may even have to adjust some of your normal entry levels in parts of the curve where you suspect there may be a large player and/or algos present. For example, the EDZ8-Z9 spread structure seems persistently bid to the curve. However, I believe this is probably a larger player bidding it (absolutely or relatively against something else), and we may have algos joining, since they know they can sell this back out fractionally higher (relative to the rest of the curve) to the Z8-Z9 bidder. We have some confirmation that this could be position-building from the larger OI in EDZ9. We have EDM8 leading selloffs going into the closing hour most days. The EDM8 OI has been roughly unchanged the past few weeks. So this feels more like some kind of spoofing.
In this environment, being patient on structures until you think things are at “crazy” levels makes sense. Fortunately for us, there are a few trades that looked “crazy” last week.