I'm excited to announce that we've launched auction imbalance data for US equities on Databento, starting with Nasdaq. We'll be rolling out the same for other US equities trading venues over the next three months. Why is this important? Most data providers source their US equities data from the CTA/UTP SIPs. In contrast, Databento sources all of its US equities data from proprietary data feeds of the exchanges. This allows us to provide richer data than the SIPs, including: Full depth of book information from trading venue Odd lot transactions Order imbalance around opening and closing auctions Prior to this release, we already provide the first two. Full depth of book information at each venue provides a more accurate picture of liquidity and allows the construction of a wide class of signals that can't be constructed with the consolidated top-of-book SIP feeds. Likewise, a significant amount of liquidity is available in the form of odd lots. The SIPs obscure this information and only display aggregated liquidity in groups of 100 shares. Odd lots are frequently priced better than round lots, providing a more meaningful view of activity inside NBBO, and the possibility to gain price improvement over NBBO. About 30% of trading, by notional volume and share count, on US equities markets takes place within the NBBO — a trend that continues to increase. Databento makes all this information more accessible and visible. Finally, the increased shift to passive investing in the US markets has resulted in a significant concentration of activity at the opening and closing auctions. Auction imbalance data allows users to have peek into activity during these times.
^ Yes, we also have NYSE, NYSE American, and NYSE Arca imbalance data since 2018, and have recently been approved for real-time distribution of all of them. We're holding off on launching these until we're close to releasing real-time data — keep in mind that the non-display fees are quite hefty for all of them, and we're working on making sure there are more accessible tiers.
any news on when the opra feed will be ready? How much historical data will be available. What is the price point and or pricing model for that feed?
For real-time OPRA data: Our usage-based fees will be about $1,500-$3,000 per month for the full feed (all trades and quotes, all symbols). Most of our users aren't consuming all symbols on the full feed. For example, we have a quant hedge fund consuming only about 10% of all CME options. So assuming that's your median use, that's about $150-300 per month. We also pass OPRA's license fees straight through. These vary: $1.25 per month per non-professional device. Potentially $600 for indirect access fee. $2,000+ per month for non-display, enterprise use. So most individual users will see all-in fees around $151-901 per month, while most enterprise users will see all-in fees around $3,500-$5,600. At this time it looks like a June release date looks most likely — we have about 13-14 exchanges we're rolling out this quarter. Unfortunately, we don't intend on providing much historical data at this time. History will probably start around this Monday, March 27, 2023. The reason is that we don't know any third-party source that can backfill OPRA data with the same standards as we have. We use hardware-based PTP timestamps and have practically zero gaps on our capture.
Yes it will be, and we only bill a user for the amount of data that they consume, so it will be much cheaper to subscribe to specific option chains than the full full feed.