By Jason Zilberbrand, President, VREF
A blog post from one of the new AI valuation tools has been making the rounds.
Its claim: “legacy” valuation guides price engine time with a straight line, the market doesn’t behave that way, and therefore the old math — and by implication, the people behind it — should be retired.
It’s a well-written piece.
The author has a genuine talent for marketing. But before anyone reprices an airplane based on it, let’s walk through what the analysis actually measured, because the details matter — and in valuation, the details are the entire job.
What they measured — and what they didn’t
The study fit six curve shapes to listing prices — asking prices — across 285 piston and turboprop models. Read that again: asking prices.
Anyone who has closed an aircraft transaction knows the distance between ask and close. An asking price measures a seller’s hope, a broker’s strategy, and a market’s mood. It does not measure what a buyer wired into escrow. Fit a curve to ten thousand asks and you have built a beautiful map of seller psychology — not a market curve.
VREF’s data is built on actual transactions, dealer-reported sales, and three decades of market history. That distinction isn’t a technicality. It’s the difference between what people wish and what people paid.
The referee is also a contestant
The post’s headline number is a “$26,367 average gap” between flat engine-hour math and the market. But look at what that gap is measured against: their own fitted curve. The analysis assumes its model is the truth, then scores the straight line against it.
That’s not an error rate. That’s the distance between two opinions, one of which graded the contest.
The honest test of any valuation model is out-of-sample performance against real sale prices: predict the number before the deal closes, then check. That test is absent from the post. Meanwhile, buried in their own results is the average signed error of the straight line: +$2,566 — on aircraft worth hundreds of thousands of dollars, very close to zero. By their own arithmetic, the method they’re retiring is nearly unbiased.
A contest designed for one winner
The post’s winning curve — a power law — was selected by letting the model search a grid of exponents and keep whichever scored best. Give one contestant six shapes and a tunable dial, give the other a single straight line, run it on small per-model samples of noisy asking prices, and the flexible model wins almost every time. That’s not a discovery about the aircraft market. That’s a property of curve fitting.
The straw man in the hangar
Here’s the part that matters most to anyone who actually buys, sells, finances, or insures aircraft: nobody competent applies flat dollars-per-hour blindly from overhaul to run-out. Not VREF’s methodology, and not any accredited appraiser working today.
The fresh-overhaul premium is real. The run-out discount is real. The mid-time plateau is real. Appraisers have priced these for decades — alongside the variables no scraped listing will ever surface: logbook quality, damage history, corrosion, engine program enrollment, prop and avionics status, missing records, and the difference between an overhaul performed by a reputable shop and one performed by a signature.
The straight-line-per-hour figure in a valuation guide is a reference point inside a larger methodology — the starting point of an adjustment process, not the end of one. Discovering that the market is nonlinear and announcing it as news is like discovering that airplanes need fuel. Every professional in this industry already prices that curve. It’s why appraisals exist.
The question to ask any free valuation
There’s an old rule that applies to valuation tools the same way it applies to everything else online: if the product is free, ask who’s paying for it.
Owners on type-club and pilot forums have noticed two things about instant AI valuations. The numbers often come in flattering — higher than owners themselves expected. And every report seems to end with an offer: want to be connected to a broker or an insurance agent?
A valuation that flatters you is a wonderful customer experience. Delighted sellers share their reports. The flywheel spins. But a flattering number isn’t a defensible number — and there is no faster way to lose money in this industry than to price an aircraft on a number built to make you feel good.
So before you rely on any valuation — ours included — ask three questions:
1. What data is underneath it? Transactions, or scraped asking prices?
2. Who signs it? An accredited appraiser whose work survives a credit committee, an insurance dispute, an IRS review, or a courtroom — or an algorithm with a disclaimer?
3. What does the publisher earn from your number? A subscription fee is a clean incentive. A referral commission is not.
Where the number has to hold up
When a lender takes an aircraft as collateral, when an insurer writes a hull policy, when an estate gets settled or a partnership dissolves, the valuation gets stress-tested by people with money at risk and lawyers on retainer. That is the standard VREF has been held to for more than thirty years, and it’s the standard we built VREF Online to meet: transaction-grounded data, accredited appraisal methodology, and values we are willing to defend — in writing, under our own name.
Innovation in valuation is welcome. Genuinely. More data and better tools make the whole market healthier. But innovation starts with measuring the right thing — and the right thing has never been the asking price.
When the number has to hold up, VREF it.
VREF has provided aircraft values and accredited appraisals to lenders, insurers, legal professionals, and owners for over 30 years.
For accurate, defensible aircraft valuations trusted by lenders, insurers, attorneys, operators, and aviation professionals worldwide, get started with your VREF Online Membership today.
Fly safe. Stay smart.
