Opening Quote: First Tick vs 9:30:05

The bot form's Opening Quote setting offers two definitions of "the open": the first tick of the day, or the price at 9:30:05. They sound interchangeable. On a calm morning they nearly are. On a gap morning they can differ by enough to flip an entry filter. This is the white paper on why the two values differ, how big the difference really is across three years of SPX tick data, and how to choose.

What the setting controls

Opening Quote lives in the Trade Conditions section of the bot form, alongside the entry filters it feeds (see Bot Settings: Trade Conditions and Entry Filters). It decides which price counts as the underlying's opening quote for the day:

• First Tick: the official opening price your broker reports for the day. tradeSteward reads the day-open value from the broker's quote data. For an index, that is the first value the index publisher put out after the bell.
• 9:30:05: the first index quote at or after 9:30:05 AM Eastern, taken as the midpoint between the bid and ask. If no quote lands during that second, the next quote after it is used.

Whichever you choose applies everywhere the bot references the open, and that is a longer list than the entry filters alone:

• The Overnight Gap filter (yesterday's close to today's open), the Intraday Change filter (today's open to now), and any other entry filter built on the opening value.
• Strike selection, whenever a leg targets its strike in Points or Percent ITM/OTM from Open. The opening quote is the base those targets measure from, so changing this setting can change which strikes the bot selects, not just whether it trades.
• The Opening Range Breakout filter's opening range. On First Tick the range starts at 9:30:00; on 9:30:05 it starts at 9:30:05. This one follows the setting for every symbol, including VIX.
• The opening quote recorded on the bot's trade.

One bot, one definition, used consistently all day.

The measurements in this article are from SPX-class cash indices, but the setting acts on other symbols too. For a single stock or ETF such as SPY or QQQ, First Tick reads the opening price the broker reports for the day, while 9:30:05 reads the first recorded quote of the session, taken as the midpoint between the bid and ask. Those are two different sources, so the two modes can produce different values for an equity as well; this article does not measure how far apart they typically sit. NDX and RUT follow the toggle like any other symbol: on First Tick they use the broker's reported day open. On 9:30:05, because neither reliably publishes a fresh value by 9:30:05, tradeSteward instead waits for the first quote that moves off the pre-bell value, with a fallback a few minutes into the session. VIX is the exception: the toggle does nothing to its opening quote, and a VIX bot always uses that same wait-for-first-movement method. Even a VIX bot feels the setting in one place, though: the Opening Range Breakout range start noted above. For a cash index like SPX, the two definitions genuinely differ for a structural reason, and the rest of this article is about why.

Why an index's first tick is not the open

A cash index has no trades of its own. SPX is a computed value: the index publisher continuously recalculates it from the prices of its component stocks. That computation has a well-known quirk at 9:30. Until a component stock prints its first trade of the session, the index calculation carries that component at its previous closing price. The very first index tick of the day is therefore a hybrid: a handful of components at this morning's opening prices, and the rest still frozen at yesterday's close.

Over the next few seconds, components complete their opening auctions and tick in, and each one that does pulls the index toward its true market level. The first tick is a snapshot taken during that convergence, not after it. This is the standard account of how a cash index opens, and the tick data below behaves exactly the way it predicts: the first tick sits systematically closer to yesterday's close than the settled open, and the bigger the overnight move, the further it has left to travel.

The timing is remarkably consistent. Across 751 trading days of SPX time-and-sales data (September 2023 through August 2026), SPX never printed a value during the 9:30:00 second. The first tick arrived during the 9:30:01 second on 97.7% of days, during 9:30:02 on 2.0%, and during 9:30:03 on the remaining 0.27%. From there SPX prints roughly once per second, so the 9:30:05 quote typically lands about four seconds and a handful of prints after the first tick.

What each choice measures: a worked example

April 7, 2025 sat in the middle of that spring's tariff selloff, and it produced the largest disagreement between the two settings in the entire three-year sample. SPX was set to open dramatically lower.

The two opening quotes ended up 44.03 points apart, 0.889%, and most of that landed in one step: the index fell 39.82 points in the single second between the first print and the next one. That is the pattern the stale-component account predicts: a first published value still anchored toward the prior close of 5073.28, repricing within a second as more components opened. The tape cannot show the index's composition directly, but it moves exactly the way the mechanism says it should.

Now watch what that does to a filter. Measured from the prior close, the First Tick opening quote showed an overnight gap of -119.49 points, -2.36%. The 9:30:05 quote showed -163.52 points, -3.22%. A bot configured to stand down when the market gaps down more than 3% trades on one setting and skips the day on the other, from the same tape, on the same morning, on exactly the kind of morning that filter exists for.

The market was not done falling, either: by 9:31 SPX stood at 4900.07, below both opening quotes and far closer to the 9:30:05 value. That is the usual outcome but not a guaranteed one; as the averages later in this article show, the 9:30:05 quote is the closer read on about two out of three mornings, not all of them.

How big the difference typically is

So on a median day the two definitions of the open sit about 0.05% apart, three to four SPX points at 2026 index levels. One day in five they differ by more than 0.10%, and a few times a year by more than a quarter percent. The record is April 7, 2025, the morning the worked example above walks through print by print; the large-move table below returns to it.

One thing the data rules out is a directional lean on ordinary days: the signed mean difference is +0.0008%, indistinguishable from zero. The first tick is not systematically high or low overall. It is systematically behind, in whichever direction the market moved overnight, which brings us to the finding that matters most in practice.

The two charts below put every one of those 751 sessions on screen: first the difference day by day, then the same days collapsed into a distribution.

A fair question about any three-year average is whether a couple of violent stretches are doing all the work. Grouping the same days by calendar quarter answers it: the difference swells in rough quarters and shrinks in calm ones, but it never goes away.

The overnight gap drives the difference

The size of the first-tick error is not random from day to day. It tracks the overnight gap, and it tracks it tightly. Regressing the signed divergence against the signed overnight gap across all 751 days gives a slope of 0.152 with a correlation of 0.81 (R² = 0.66, slope 95% confidence interval 0.144 to 0.160). In plain terms: on gap mornings of a quarter percent or more, the first tick misses about 15% of the overnight move, whichever direction the move went. The conventional stale-component account attributes that to the slice of the index still carried at yesterday's close when the first value is published; the regression measures the size of the miss, not its composition.

A note on how the gap itself is measured. The statistics in this section run the overnight gap from the prior session's close to a settled index value at 9:31, after the open has fully formed, rather than to the first tick. (The large-move table further down is the deliberate exception: it shows the gap each setting itself would have reported.) The first tick is the very price whose accuracy is being tested, so measuring the overnight move against it would build the error being studied into the yardstick. The 9:31 reference keeps the yardstick independent, and re-running the analysis against the 9:30:05 price instead gives figures that agree to well within the margin of error.

Above that quarter-percent line, the 15% fraction is stable: bucketing days by gap size, the ratio of divergence to gap stays in the same 0.12 to 0.18 band, with no trend either way, from quarter-percent gaps up through the largest gaps in the sample. Below it the ratio runs much higher (the median is about half in the smallest-gap bucket), but that is arithmetic rather than signal: dividing an ordinary few-hundredths-of-a-percent settling move by a tiny gap produces a large ratio while the absolute difference stays negligible. Nearly half the sample (355 of 751 days) gapped less than a quarter percent, and on those days the two settings typically sat only 0.03% apart.

The figures in the next table run a little above 15% of the gap because averaging absolute differences picks up ordinary opening noise along with the gap-driven part; the slope describes the signed relationship, while the table describes what a typical morning of each size looks like. The 2.00% row rests on just seven mornings, so read it as indicative.

The direction is as predictable as the size. On the 445 gap-up mornings in the sample, the 9:30:05 price extended further above the first tick 73.0% of the time, by a mean of +0.047%. On the 306 gap-down mornings, it extended further down 78.4% of the time, by a mean of -0.067%. The fitted slopes for up gaps and down gaps (0.156 and 0.166) are not distinguishable in this sample.

Here is what the choice looks like on the biggest mornings.

The fourth row is the worked-example morning from earlier, and the table shows it was extreme but not alone: every morning here is the same story at a different scale. The miss also runs one way: on all ten of these mornings the first tick reported a smaller gap than the 9:30:05 quote, never a larger one. Across the full sample, the first tick understates the size of the overnight gap on 554 of 751 days (73.8%), and on the 58 days that gapped 1% or more by the 9:30:05 measure it understates on 56 of them (96.6%).

Two mornings keep the table honest. January 27, 2025 gapped more than 2%, yet the two settings differed by only 4 points (0.07%): a large gap raises the odds of a large disagreement without guaranteeing one. The mirror image came on March 13, 2026, a day too small to make this table: the first tick showed a 0.02% gap, an essentially flat open, while the 9:30:05 quote showed 0.50%. A gap filter reading the first tick saw no overnight move at all on a morning that actually gapped half a percent.

The same logic applies to Intraday Change filters, in mirror image. The first tick sits closer to yesterday's close, so a First Tick bot starts measuring intraday change from a point partway through the overnight move, and the unfinished tail of the gap gets counted as intraday movement instead. On a typical 1% gap-up morning that adds roughly +0.13% to every intraday change reading; on a gap-down morning it subtracts the same. The offset is set in the first seconds of the session and shifts every open-referenced reading for the rest of the day.

Both findings are visible in the raw days. Every session in the sample appears in the scatter below, and the bars underneath regroup the same days by gap size.

Second by second: watching the index settle

Following the tape out from the first tick shows where the convergence actually happens.

The story is one huge step and then a taper. The move from the first tick to the next second averages 0.064%, and 0.137% on large-gap days, which is most of the total divergence on its own. The move from :02 to :03 is a fifth of that, and from there each second contributes around 0.01%, decaying slowly to about 0.007% per second by 9:30:15. The elbow is at 9:30:03, and by the 9:30:05 sample gap capture has flattened out near 99%.

This is why the sample is taken at 9:30:05. It reads the index after the convergence has happened rather than during it, with margin to spare: on the roughly 2.3% of days whose first print does not arrive until 9:30:02 or 9:30:03, a 9:30:05 sample still leaves several prints of settling behind it. The first tick, by contrast, is taken at the one moment in the session when the index is furthest from where it is about to sit.

It is equally honest to say that 9:30:05 is not "the true open." It is a sample taken shortly after convergence rather than during it. Measured against a 9:35 reference instead of 9:31, the 9:30:05 price reflects 96.5% of the move, because the market itself keeps moving in those minutes. There is no instant at which an index open is final; 9:30:05 is simply a moment by which the stale-component effect is spent and ordinary trading is what remains.

Charted, the convergence and its taper look like this:

Choosing a setting

9:30:05 is the better default when your filters mean to compare against a settled opening value. Overnight Gap and Intraday Change thresholds almost always intend that: "skip the day if the market gapped more than X" is a question about where the market opened, not about the first published print. Direct measurement backs this up: across the sample, the first tick sat an average of 0.090% from the settled 9:31 price while the 9:30:05 quote sat 0.050% away, and 9:30:05 was the closer of the two on 68.4% of days. It is also the steadier base for comparing one day's behavior to another's, since it reduces a source of variation that has nothing to do with the market and everything to do with auction timing. Here is that comparison across every morning in the sample:

First Tick is a deliberate choice, not a mistake. It is the earliest read the market offers, and if your strategy trades the open's first move, an entry decision at 9:30:10 may prefer the value from nine seconds ago over waiting on the settled one. It is also the right choice when you are matching a data source, a historical study, or another platform that defines the open as the first print, where consistency with the reference matters more than the print's composition.

Remember the scope of the measurements: they are from SPX-class index bots. The setting still acts elsewhere. A stock or ETF bot reads the broker's reported open on First Tick and the first recorded quote of the session on 9:30:05, two different sources whose gap this analysis does not measure. An NDX or RUT bot reads the broker's reported open on First Tick, and the wait-for-first-movement detection described earlier on 9:30:05. A VIX bot uses that first-movement detection for its opening quote regardless of the toggle, though an Opening Range Breakout filter's range start follows the setting even on VIX. And a leg that targets its strike in Points or Percent ITM/OTM from Open selects strikes off whichever opening value the setting chooses.

Whichever you run, two things matter more than the choice itself: know which definition your thresholds were tuned against, and read the matching value when you are checking a filter's work. If you switch a bot's setting, revisit any gap or intraday thresholds tuned under the old one; on gap mornings the ~15% offset is the difference between a filter that trips and one that does not.

See both values live

You do not have to take any of this on faith. The Bot Filter Values page shows, for indices like SPX, the First Tick and 9:30:05 as Open boxes side by side, each with its own opening quote, intraday change, and overnight gap for today's session. On a gap morning, open it and watch the two boxes disagree in exactly the proportions described above. The page is documented in the Bot Filter Values guide; when a filter decision surprises you, read the box that matches the bot's Opening Quote setting before concluding anything misbehaved.