Two Year Fat Finger

sharkie · Post by **sharkie** » Fri May 29, 2009 1:37 am

Around 7:30 a.m. CST today the Sept. two year treasury future [TUU9] looked like it traded down >100 ticks executing all resting bids all the way down [several thousand contracts] and it was immediately bid back. I don't think this trade was busted, does anyone know what caused this? The June future only went down 5-10 ticks and the cash didn't budge either. I can't imagine the feeling of vaporizing several million dollars in a millisecond like that or the rejoice of accidentally leaving in a thousand lot bid a point down. Any comments on how this happened would be appreciated.

Thanks,

Sharkie

Martinghoul · Post by **Martinghoul** » Fri May 29, 2009 10:03 am

It was a fat finger, apparently, but CBOT let all trades stand. I don't quite know what other insight is to be had. Things like this have occurred before (bund a short while ago) and will occur in the future...

knocks_rocks · Post by **knocks_rocks** » Fri May 29, 2009 5:07 pm

TU is pit contract no?

Edit: Fat finger on a pit trade seems unlikely, electronic maybe. Perhaps just an error?

sharkie · Post by **sharkie** » Sun May 31, 2009 7:17 am

"Things like this have occurred before (bund a short while ago) and will occur in the future..."

I agree these things are rare and I am trying to imagine a world where you take advantage of black swans...

Is stacking orders in size off of some rudimentary valuation not an edge producing model if mullti-million dollar opportunities present themselves every half-year or less? The cost to place orders is basically 0, so the profit opportunity is -only- based on a successful trade. We aren't talking about something meaningful to a large fund prop desk but to a small desk the lean cost/profit margin produces a worthwhile model to fish for the bone.

I enjoy all the discussion on this forum and only mean the best.

-Sharkie

ax · Post by ax » Sun May 31, 2009 7:13 pm

i would think the people running automated spreaders were able to take advantage of this

sharkie · Post by **sharkie** » Sun May 31, 2009 7:23 pm

For sure, but for the most part people running say an automated spreader against eurodollars or the cash or whatever, probably weren't working to leg into something 200 ticks better than the current market with limit orders.

They may have had some backups for 5-10 ticks off the market. Especially if this was effectively a market order you would have to be in there 150 ticks off the market in order to really benefit, because it's executed and gone, it's not like you have a few milliseconds to just execute the trade against resting liquidity that far down and coincidentally get price improvement on the autospreader level you were trying to work.

These are only my guesses actually, so take it for what it's worth.

Martinghoul · Post by **Martinghoul** » Mon Jun 01, 2009 9:55 am

Sharkie, I wasn't being dismissive, so apologies if I sounded like it...

I think most people who had orders in various spread books (the most obvious being the 2y roll) weren't particularly happy about this, as it's volatility that they're not getting paid for. The other question for these "opportunities" is volume. This one was somewhat uncharacteristic in that it was a VERY large move (2pts on 2y note is punchy) and relatively large volume. In most cases (again, like the famous bund last year and short sterling a couple of months earlier), there's quite a lot less juice in it...

As I understand it (again from the bund case), this really hurts the smaller locals a lot, 'cause it normally triggers their stops and they're not able to cover in time...

autokor · Post by **autokor** » Mon Jun 01, 2009 7:36 pm

the problem with such a "fat finger" fishing limitorder is that it could be that there is a reason for the massive move . And if there is a reason you will be always filled on the false side.