Cumulative delta is NOT a measure of directional bias
How this popular measure can lead traders towards erroneous conclusions.
Cumulative delta is a widely followed metric by traders, particularly within the narrower timeframes (i.e., intraday). And, admittedly, it can be an informative indicator for grasping broader market context … if it is correctly interpreted. However, many traders seem to misunderstand the information content that delta is actually conveying, often conflating the direction of the delta with the direction (or anticipated direction) of the price action. In this brief post, we hope to explore the meaning of delta and how to limit drawing erroneous conclusions from its trajectory.
Cumulative Delta = Volume executed at Ask Price - Volume executed at Bid Price
First, let us properly define delta. In orderflow parlance — not to be mistaken for the risk factor denoted by the greek letter delta [Δ] that is widely used in options — delta is a measure of the net differential between those executing buy market orders on the ask (or offer price) and those initiating sell market orders on the bid. In trading jargon, the former is commonly referred to as lifting the offer and the latter as hitting the bid. Market orders are widely interpreted as a measure of aggressiveness by traders because they ensure a fill in return for sacrificing (or paying the cost of) the bid-ask spread. Prioritizing a fill at the expense of electing to use a passive limit order — which gurantees a given price (or better), but incurs the risk of not completing the order — is usually taken as an indication of greater urgency, need, or conviction.
Delta simply describes how traders have chosen to execute their orders, either passively or aggressively.
However, this is where paying sufficient attention to the definition of delta is important: delta simply describes how traders have chosen to execute their orders, either passively or aggressively (i.e., limit or market orders, respectively). By itself, delta tells one very little about the directional tendencies or biases of the participants. A trader can be bullish and still choose to wait patiently for fills on the bid; conversely, bearish and elect to execute their orders passively on the offer price. Even if their bias is aggressively tilted towards a particular directional posture, a trader does not need to execute their orders aggressively at market. And this takes on added significance in derivatives because such instruments are widely used as hedging vehicles and may not represent the underlying speculative position. As such, price insensitivity on the hedge is not uncommon, and transactions possess the potential to mask a participant’s true positioning or directional bias. And keep in mind that delta can be measured over any periodicity, but typically an entire session’s delta is referred to as cumulative delta.
Cumulative delta surge into the closing bell driven by size participants (S&P 500 futures).
Observations about how traders choose to execute their orders takes on added significance in derivatives markets because such instruments are widely used as hedging vehicles and may not be representative of the underlying speculative position. As such, price insensitivity on the hedge is not uncommon, and transactions have the potential to mask a participant’s true positioning or directional bias.
Furthermore, in some markets, the relative level of liquidity and volatility will often directly influence the way a trader decides to execute their orders. For example, a trader involved in a market that is fast-moving and covering a lot of ground — hence, more likely to reach a preferred price — might be less inclined to force an aggressive entry or exit and more liable to let the market come to their preferential location. On the contrary, perhaps a similar highly volatile environment would encourage a different trader to pursue hard (passive limit order) stops for their exits because they do not want to incur the adverse slippage that is common in fast (less liquid) markets; while in slower markets this same trader prefers a go-to-market (aggressive) approach with their orders because of the low likelihood of sacrificing much ground in terms of price. As one can easily see, none of these decisions are necessarily driven by the participant’s directional bias, and are primarily influenced by the mechanics of the underlying market.
An example of how cumulative delta can be misleading
If this seems a bit confusing, perhaps a simple example will better illustrate the concept. Let us assume that there is a market comprised of 11 participants: one very large trader and ten smaller ones. We’ll assume that this single large trader is looking to buy 1,000 contracts but does not need a fill at a single price and can afford to be a bit patient in the hope of gaining more favorable execution. The ten smaller traders are looking to sell 100 contracts, either immediately or slightly above market. Let us further assume that the instrument in question is priced at $100 and is tradeable in increments of $1.
At the opening $100/101 (bid/ask), the large buyer sits on the bid with his order. The first seller steps in and hits the bid (executes a market sell order) for 100 contracts and is filled immediately. The large buyer has gotten 100 contracts of his 1,000 lot order filled. Cumulative delta is now -100.
The next seller steps in and hits the bid getting a fill. Delta is now -200.
The next seller steps in and hits the bid getting a fill. Delta is now -300.
The following seven sellers all look at themselves and say, "There's buy-side demand for this. I'm raising my price to $102”. The offer price then shifts up to $102, and the large buyer says, "ok, I'll bid $101 now”, so we end up with a new 101/102 inside-market.
The 4th seller hits the bid at 101. Delta is now -400.
The 5th seller hits the bid. Delta is now -500.
The following five sellers say to themselves, "Hold on now, there's some real buying demand. I'm raising my price to $103”. The large buyer then says, "Ok, I'll bid $102". Our inside-market moves to 102/103.
The 6th seller says, "Wow, I can get $102 for certain execution now and just a few minutes ago I had to settle for only $100… I'll take it!" He hits the bid. Cumulative delta is now -600.
The large buyer is now getting a little anxious for the complete fill because he can see that the price keeps increasing. He decides to buy 300 contracts at the offer ($103), but since there are not more than 100 lots at each price, he ends up sweeping the 103-104-105 prices with that market order. Delta is now -300, and the bid/ask inside-market has moved to 105/106.
The last seller decides to just hit the bid now at $105, and the buyer takes his last fill of 100 contracts to complete his 1,000 lot order.
Cumulative delta is -400.
So was this price action bearish? Of course not! Price is higher, and yet cumulative delta trended lower the entire time. The mechanics of that evolution trace directly back to the collection of participants and their intent (or needs) at the time. If a trader happened to look at the cumulative delta when it trended down, hitting a nadir of -600 (with a 102/103 inside market), and then interpreted that as a sign of bearishness — choosing to short the market in question — that trader would have been subsequently caught up in the sweep higher that occurred across the 103-105 price advance and possibly had themselves stopped out in the process.
As with most things in markets, context matters!
Of course, this is a highly simplified example, and there are many ways in real markets to arrive at similar (or different) outcomes. And, crucially, we still don't know much about the large buyer’s intent. That buyer could have been purchasing in size in order to hedge an even larger short (core) position elsewhere. But hopefully this helps to illustrate how delta can diverge from price without any real net bearishness (or bullishness) in the price action itself. As with most things in markets, context matters!