Order Flow, Footprints, and Volume Profiles in rates products
This is a topic I've been thinking about for quite a while, and would love to hear other rates-traders' perspectives and observations.
In terms of overall DV01 going through the rates market every day, the vast majority of trading occurs in the cash / otc markets (treasuries, swaps, and other interest rates derivatives), leaving no visible footprint for futures traders to analyze. Moreover, a significant portion of futures volumes are not outright position trades, but rather hedges of cash products or single legs of curve or basis trades, rendering their individual footprints much less meaningful to an order-flow trader. We might see a 200-lot print in ZB but never get to see the other side of the trade going through the cash market a few seconds later, leaving us no clue as to whether trade represents naked or hedged interest-rate risk.
All of this suggests to me that trading systems which rely heavily on volume profiles, cumulative delta, or other order flow data would be disadvantaged by the information asymmetries that exist in futures markets (this would probably apply in currencies as well), but there are still quite a few traders who employ successful strategies using these limited data points.
Perhaps this is because futures markets capture a lot of the short-term (intra-day) volumes, while cash and derivatives markets tend to be dominated by longer-term players, i.e. traders operating on daily / weekly / monthly time horizons. Nonetheless, the large hedging flows that go through futures still print a rather one-sided picture.
So my questions for other interest rate futures traders is: How do you account for the hidden flows going through cash and OTC derivatives markets? Have you found an effective way to filter out some of the "noise", or data-skew arising from cross-market hedges and paired-off basis or curve trades? Is there any measurable benefit to aggregating volumes across the rates futures complex to better approximate the outright DV01 going through in futures, and if so, are there any off-the-shelf products that already do this?
Looking forward to everyone's thoughts.
The following 2 users say Thank You to Bladesmith for this post:
OK I'll try to answer my own questions, but please bear in mind that since I am not currently employing any of these strategies in rates products I can only try to provide my best guess here - a theoretical exercise that I hope might provoke some further discussion
Curve trades can be filtered out simply by watching the orderflow across products. Duration ratios are easy enough to calculate and a simple spreadsheet formula can allow any rates trader to follow aggregate duration-adjusted cumulative delta for the overall treasury complex.
Since there's not a ton of cross-pollination between the treasury and the eurodollar folks it's probably not necessary to incorporate those "footprints" into ones order-flow analysis, especially when focused on the longer end instruments.
Meanwhile, with cash treasuries and swaps markets being more of an "investors market" and futures being more of a "traders market," intra-day directional flows are probably much less meaningful in cash. Your typical swaps trader isn't exactly going to bail on a position if it moves half a basis point against him in two minutes, and while there are still quite a lot of fast-money accounts operating in the cash space, their time-horizons are still typically a bit longer than your typical scalper. Holding positions overnight is not at all uncommon for these traders.
Finally, cross-market hedges simply don't represent enough volume or directional bias to matter all that much, as overall flows are still dominated by directional short-term traders.
Therefore, with a bit of simple math and a trained eye, an experienced rates trader can do a pretty good job analyzing footprints and order flow, despite all the other volumes going through the various otc rates products.
Again, all of the above is only a best guess. Any other experiences, observations or perspectives - especially those contradicting my own loosely held assumptions - would be very much appreciated.
The following 2 users say Thank You to Bladesmith for this post:
Thanks for opening up this discussion. As a neophyte ZB trader using volume profile/order flow(Jigsaw Platform) as part of my trade criteria, getting an accurate volume profile across all active rates markets would be essential (I would assume). I can't really add any insights to the discussion yet (currently wrapping my head around basis/curve trades ;-) but look forward to others adding to this thread. I wonder if the folks at Jigsaw would have any comments?
The following user says Thank You to Vantrader66 for this post:
Agreed - an interesting question, Bladesmith. I've wondered about this in regard to the exchange-traded options (stocks and futures) vs their respective OTC activity. Even if an analyst is able to figure some *aggregate* sentiment of price/volatility/etc (e.g. plotting the IV surface and looking for reversion/continuation of higher moments) from the 'lit' data in options, the OTC 'dark' data could dwarf the 'lit' in pure notional value and/or be heavily skewed toward the opposite sentiment.
The only pondering I've kept in mind for dealing with this issue is to look at market behavior near an exchange's open & close, esp when daytrading - in the case of futures, perhaps you relax your model forecast(s) for the near-term & avoid trading around those times (e.g. Europe close), *if* elevated price volatility or strong reversal are commonly observed around those times (ditto for macroeconomic data, of course).
You've read my mind. Was thinking about this the other day in relation to the bund. The futures market is simply a reflection of the cash bund market, and the order book for that is not available to us. And the futured market is a derivative of this market. The only thing I can say is that there are some very big and very clever traders in the bund/ZN/ZB markets, so essentially we're following in their footsteps.
Have to say that I don't think the currency futures markets are up to much as their volume is seriously light, and they are the derivative product of a huge market where trades happen OTC and all kinds, and there is no centralised exchange.
How do we know what the principle cash markets are doing in terms of volume? I don't think we ever do. The only thing we can do is pick the most liquid and thick markets where the smartest traders are.
How I'd love to have a volume profile chart with volume attached to the cash/spot /treasury market for the bund.