Datasets:
You need to agree to share your Hugging Face information to access this dataset
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
You agree not to redistribute this data.
Log in or Sign Up to review the conditions and access this dataset content.
Futures Yearly Front Month Cost Params
Measured trading-cost parameters for futures, one row per contract root and year -- 329 roots over 2017-2026.
Columns
symbol: the contract root (CL,ES,GC), not a contract. This is what a continuous series trades.year: calendar year the parameters were measured over.price: the mean mid-quote of the front month.spread_bps: the full quoted spread of the front month in basis points, time-weighted over the quotes of the sampled sessions. A round trip crosses half of it on each side.lambda_bps: null here. Kyle's slope needs signed trade flow, which lives in the trade files and not in the quotes this table reads. Use the per-contract dataset for it.
Why it exists
A table keyed by contract cannot price the front month, and not merely because
its median is wrong: its minimum across a root's contracts is unrelated too.
A contract has a lifecycle, so its yearly row averages the months it was
far-dated and wide with the months it was the front month and tight. Measured
on 2020-08-05, the active gold contract quotes at 0.81 bp, while across
every GC contract of 2020 the minimum reads 7.92 bp and the median
39.90 bp -- factors of ten and fifty, in the same direction every time.
So the front month is selected per session: the busiest contract of the session is the front month, by definition rather than by a rule.
How to check it, for free
A spread cannot be tighter than one tick, so tick / price is a floor that
needs no measurement at all. The ten most liquid roots land on it:
| root | spread, in ticks | range over 2017-2026 |
|---|---|---|
ZN |
0.99 | 0.98 - 1.00 |
ZB |
1.00 | 0.98 - 1.02 |
ES |
1.04 | 0.98 - 1.14 |
ZC |
0.91 | 0.86 - 1.10 |
It is a floor test, not a band. GC reads 5.40 ticks in 2026 only because
gold trades at $4,754 against $1,264 in 2017, while its spread in basis points
is flat across the decade (0.68 to 1.14 bp). A ratio above one is not
suspicious; one below ~0.9 is.
Caveats
- Sampled, not exhaustive: one Wednesday per quarter, 38 sessions over ten years. A Wednesday is never a market holiday. This measures the level of a root's spread, not its day-to-day variation.
- A session is dropped when it carries too few quotes or too few roots, which is how a truncated download is caught rather than averaged in -- three November 2017 sessions were dropped on that test.
- Rows with no spread, or a zero spread, are not published. The second happens on products whose price is a differential near zero, where a relative spread has no meaning and a zero would let a backtest trade that root for free.
- Downloads last month
- 3,221