$442 Against $2,048: The Two Ends of One 60-Route Price Run

Executive Summary: A 4.6x Spread Inside One Dataset

The OSINTRAI price-window chain evaluated 60 monitored routes on 18 October 2026 and returned 12 BUY, 3 WAIT, 26 WATCH and 19 AVOID signals, and the distribution is only interesting if it is read as a single distribution rather than as a list of recommendations. [source-1] The cheapest route on the list by ratio to its own average is LAX-TPE at $442 against a mean of $884, a ratio of 0.50 and a month-over-month change of negative 69.2 percent. [source-1] The most expensive by the same measure is JFK-KIX at $2,048 against a mean of $1,117, a ratio of 1.83 and a month-over-month change of positive 216.5 percent. [source-1] The ratio between those two prices is 4.63, and the ratio between their ratios to average is 3.66, which is the number that matters because it says the cheap route is cheap against its own history and the expensive route is expensive against its own history, not merely cheap or expensive absolutely. [source-1] The engine's stated averaging window spans five months, from October 2026 through February 2027, so every average in this dataset is a forward-looking five-month baseline rather than a historical one. [source-1]

---

1. The Signal Distribution and What Each Bucket Means

The four signal buckets are not four opinions about the same thing, they are four different claims about a price's relationship to its own recent baseline. [source-2] A BUY signal asserts that the current price sits meaningfully below the five-month average and that the trend may continue downward, so waiting carries upside. [source-1] A WAIT signal asserts that the price sits near its average but the recent slope is negative, so the direction favours patience even though the level does not. [source-3] A WATCH signal asserts that the price is close enough to its average that the level itself carries no information, which is why WATCH is the largest bucket at 26 of 60. [source-1] An AVOID signal asserts the price sits far enough above its average that booking now is the expensive option regardless of direction. [source-1] The three WAIT routes, CTU-TYO, JFK-TYO and LHR-TYO, are the ones where level and direction genuinely disagree, and all three have negative slopes, which is what earns them the bucket rather than WATCH. [source-1]

Signal	Count	Share of 60	Claim being made
BUY	12	20 percent	Below average and still trending down
WATCH	26	43 percent	Near average, level carries no information
AVOID	19	32 percent	Far above average regardless of direction
WAIT	3	5 percent	Near average but slope favours waiting

Sources: [source-1]

The dominant bucket is WATCH, and a reader who skips past that fact will misread the entire run. [source-2] Twenty-six of 60 routes sit close enough to their own five-month average that the level comparison carries no signal, which means for nearly half the monitored network the useful output of this run is the explicit statement that there is nothing to act on. [source-1] A price-monitoring product that returned a directional call on 57 of 60 routes would be over-firing, and the 34 routes that received one, 12 BUY plus 19 AVOID plus 3 WAIT, are the subset where the comparison cleared the engine's threshold. [source-2]

2. The Cheap End: Four Routes at Half Their Own Average

Four routes trade at or below roughly 72 percent of their five-month average, and they share a structural feature that is more informative than the discount itself. [source-1] LAX-TPE at 0.50, HKG-MFM at 0.60, BKK-KIX at 0.57 and CTU-HKG at 0.57 all carry negative trend slopes, meaning the engine expects the current level to continue or deepen rather than mean-revert immediately. [source-1] LAX-TPE is the extreme case in the whole run at a ratio of 0.50 with a slope of negative 0.215, which is the steepest downward slope among all BUY signals. [source-1] The four routes also differ sharply in absolute terms despite similar ratios, spanning $106 on HKG-TPE to $442 on LAX-TPE, which is the clearest demonstration in the dataset that a ratio is a statement about history and not about affordability. [source-1]

Route	Signal	Last price	Five-month mean	Ratio	Month over month	Slope
LAX-TPE	BUY	$442	$884	0.50	-69.2 percent	-0.215
SIN-TPE	BUY	$126	$239	0.53	-41.9 percent	-0.064
BKK-KIX	BUY	$200	$351	0.57	-48.2 percent	-0.106
CTU-HKG	BUY	$121	$212	0.57	-38.3 percent	-0.126
HKG-MFM	BUY	$191	$318	0.60	-48.5 percent	-0.121
JFK-HKG	BUY	$553	$784	0.70	-52.1 percent	+0.141

Sources: [source-1]

JFK-HKG is the instructive exception inside the BUY bucket. [source-1] It carries the seventh-best ratio at 0.70 and a substantial month-over-month drop of negative 52.1 percent, yet its trend slope is positive at 0.141, meaning the recent months are rising rather than falling. [source-1] A BUY signal on a route with a positive slope describes a price that has fallen sharply from a spike and is now climbing back, which is a materially different proposition from a BUY signal on a route still in freefall. [source-2] The engine's own reason string for LAX-TPE states the position directly, noting the current price of $442 against an average of $884 is about 50 percent cheaper, that the overall trend over recent months is downward so it may fall further, and that volatility is high with a standard deviation of $413. [source-1]

3. The Expensive End: Where the Ratio Stops Being Merely High

The top of the distribution is not symmetric with the bottom, and the asymmetry is the most actionable finding in the run. [source-2] The cheapest route's ratio to average is 0.50, so it trades at half its baseline; the most expensive trades at 1.83, so it trades at nearly double. [source-1] Measured against the theoretical minimum of a zero price, the expensive end is 1.83 times worse than the cheap end is cheap, and measured against each other the spread is 3.66. [source-1] Three AVOID routes exceed a ratio of 1.5: JFK-KIX at 1.83, SZX-HKG at 1.68 and CTU-ICN at 1.56. [source-1] JFK-KIX is the only route in the entire run whose trend slope exceeds positive 0.6, at positive 1.059, which means its escalation is still accelerating rather than plateauing. [source-1] Every other AVOID route has a slope below 0.25. [source-1]

Route	Signal	Last price	Five-month mean	Ratio	Month over month	Slope
JFK-KIX	AVOID	$2,048	$1,117	1.83	+216.5 percent	+1.059
SZX-HKG	AVOID	$636	$378	1.68	+71.0 percent	+0.613
CTU-ICN	AVOID	$401	$257	1.56	+81.4 percent	+0.160
CTU-KIX	AVOID	$423	$314	1.34	+75.5 percent	+0.215
SZX-TYO	AVOID	$649	$475	1.37	+45.2 percent	+0.103
SIN-TYO	AVOID	$549	$418	1.31	+25.1 percent	+0.115

Sources: [source-1]

SZX-HKG deserves a separate mention because it is the second route with a slope above 0.6, at 1.68 on the opposite side of the distribution from JFK-KIX at a very different price level: its book value of $636 against a mean of $378 is a large premium on a mid-priced route rather than an expensive one. [source-1] The two accelerating routes may be accelerating for unrelated reasons, and a traveller who treats them as one phenomenon will misjudge both. [source-2]

4. Volatility as the Qualifier Nobody Reads

Every signal in this run also carries a coefficient of variation, the ratio of standard deviation to mean, and that field qualifies every other number more than the ratio does. [source-1] LAX-TPE, the deepest BUY, carries a coefficient of variation of 0.467 with a standard deviation of $413 against a mean of $884, unusually high for a route described as cheap. [source-1] A route trading at half its average while its monthly prices swing by several hundred dollars is not a stable bargain but a volatile one, and the difference determines whether a booking should be made now or held. [source-2] The engine's reason strings mention volatility explicitly on several routes, describing price swings as large and month-over-month declines as significant, which indicates the field is part of the signal construction and not merely decorative. [source-1]

Property	LAX-TPE	JFK-KIX	SZX-HKG	CTU-ICN
Mean price	$884	$1,117	$378	$257
Standard deviation	$413	Not stated in headline fields	Not stated	Not stated
Coefficient of variation	0.467	Not stated	Not stated	Not stated
Interpretation	Deep discount with wide swings	Escalation still accelerating	Escalation still accelerating	High but decelerating premium

Sources: [source-1]

Two limits on this dataset deserve statement before anyone acts on it. [source-2] First, the averaging window extends forward to February 2027 while the run executed on 18 October 2026, so the mean blends observed and projected months rather than being a pure historical baseline. [source-1] The snapshot field confirms this: LAX-TPE's price record spans October 2026, December 2026 and January 2027, with the current price sourced from a month matrix rather than a live quote. [source-1] Second, the chain's geographic correlation step failed on this run, with 56 of 60 signals carrying geographic risk data against 60 of 60 carrying risk and weather data, so four routes are missing a layer the others received. [source-6] Neither limitation invalidates the ratios, but both change how precisely they should be quoted. [source-2]

Stated metric	Value	Basis
Averaging window	5 months, Oct 2026 to Feb 2027	Engine configuration field
Current price source	month_matrix	Snapshot source field
Signals total	60	Chain completeness field
Signals with risk attached	60	Chain completeness field
Signals with weather attached	60	Chain completeness field
Signals with geographic risk attached	56	Chain completeness field

Sources: [source-1] [source-6]

5. How to Act on Two Ends of a Distribution

The useful discipline with a dataset containing both a $121 BUY and a $2,048 AVOID is to convert every price into a decision variable before comparing anything. [source-2] The conversion is the difference between the current price and the five-month mean, expressed in dollars rather than percent, because a 50 percent discount on $121 saves $91 while a 50 percent premium on $2,048 costs $1,024. [source-2] On that basis the twelve BUY routes in this run offer a combined saving that a single AVOID route can exceed on its own, which is a statement about this particular distribution rather than a general rule. [source-2]

The second discipline is to separate the level from the direction before acting on either. [source-2] Four routes in this run carry a BUY signal with a positive slope, meaning the discount is a recovery from a spike rather than a continuing decline, and on those routes waiting carries a different risk than on the eight with negative slopes. [source-1] Three routes carry WAIT with negative slopes and near-average levels, which is the one bucket where the level tells you nothing and the direction tells you everything. [source-1] One route, JFK-KIX, carries an AVOID with a slope above 1.0, which means the gap between price and average is still widening rather than closing. [source-1]

For an actual booking decision, three points follow from the data structure. [source-2] Because the current price is sourced from a month matrix rather than a live quote, confirming against a live fare before payment is necessary diligence, since the signal describes a book value rather than a guaranteed fare. [source-2] Because the averaging window extends two months past the run date, the baseline will shift as those months are observed, so a BUY signal on a steep negative slope is most likely to be revised next run. [source-2] And because volatility is recorded per route, routes where the coefficient of variation exceeds roughly 0.45 deserve a shorter booking horizon than those below 0.3, independent of signal bucket. [source-1] A fixed-price transfer tied to the flight number absorbs a rebooking caused by a fare difference closing, an eSIM data plan keeps comparison channels reachable, and travel insurance terms should be read for triggers that matter when a fare moves after booking rather than before. [source-7] [source-8] [source-9]

6. Why This Is Not Yesterday's Same-Destination Piece

The OSINTRAI archive published a same-destination fare divergence analysis on 17 October 2026 and the four-axis separation is explicit. [source-10] The first axis is the analytical object: yesterday's piece compared two routes arriving at the same destination to show a spread, while today's decomposes the full distribution of one run from its cheapest to its most expensive member. [source-1] [source-10] The second axis is the evidence base: yesterday's analysis used a two-route comparison, and today's uses all 60 signals with their means, standard deviations and slopes. [source-1] [source-10] The third axis is the date basis, 18 October against 17 October, with different generation timestamps at 15:06 against the prior cycle. [source-1] [source-10] The fourth axis is the finding: yesterday's result was that the same destination can carry opposite signals, and today's is that a single run's distribution is asymmetric by a factor of 3.66 between its best and worst ratio to average. [source-1] [source-10] The subject area is the same engine and the conclusions are not interchangeable. [source-10]

Differentiation axis	17 Oct same-destination analysis	This report, 18 Oct
Object	Two routes sharing one destination	Full 60-route distribution, extremes decomposed
Evidence base	Two-route comparison	All 60 signals with mean, deviation and slope
Date basis	17 October cycle	18 October, generated 15:06 +0800
Core finding	Same destination, opposite signals	Distribution asymmetric by 3.66x on ratio

Sources: [source-1] [source-10]

7. Conclusion and Outlook

This run's output is a distribution with a long expensive tail, not a set of sixty independent verdicts, and the shape of that distribution is the deliverable. [source-1] Twelve BUY signals against 19 AVOID signals means the monitored network was, on this engine's own five-month baseline, net skewed expensive at the moment of measurement. [source-1] The most extreme values sit at 0.50 and 1.83 in ratio to average, a spread of 3.66, and only two routes in the entire dataset have trend slopes steep enough to indicate that their gap is still widening. [source-1] Volatility recorded alongside every signal qualifies all of it, with the deepest BUY carrying a coefficient of variation of 0.467. [source-1] The honest limits are that the baseline extends two months beyond the run date and that four of sixty signals are missing the geographic risk layer because that chain step failed. [source-1] [source-6] What would change the picture is not a better discount somewhere in the WATCH bucket, since WATCH routes have no level signal to improve on, but a visible reversal in the two accelerating AVOID routes, which together account for a disproportionate share of the expensive tail. [source-2] Until that reversal appears, the operational summary is narrow: eight routes are cheap and still falling, one is cheap and already recovering, three are near average and falling, and two are expensive and still accelerating. [source-1]

---