An average bottle price hides the bottles that actually change in value
The wine collection value estimator asks for an average bottle value and one growth rate. For a collection of similar bottles that is fine. For a typical cellar, which is mostly everyday wine with a few special bottles, a single average quietly gives the wrong answer, and it is worth seeing by how much.
One collection, two tiers
Suppose 48 bottles of everyday wine averaging $25.00, which, like most wine bought to drink, we will assume does not appreciate at all; and 12 bottles averaging $150.00, which we will assume grow at 6% a year. Both rates are illustrations, not predictions. Over 10 years, run separately:
| Tier | Bottles | Value today | Assumed growth | Value in 10 years |
|---|---|---|---|---|
| Everyday | 48 | $1,200.00 | 0% | $1,200.00 |
| Fine | 12 | $1,800.00 | 6% | $3,223.53 |
| Total | 60 | $3,000.00 | $4,423.53 |
The same bottles as one average
Run it once instead, and you need one average value and one rate. The average value is easy: $3,000.00 over 60 bottles is $50.00. The rate is where it goes wrong, because there are two ways to average it, and both are wrong.
| Approach | Rate used | Value in 10 years | Short of the two-tier figure by |
|---|---|---|---|
| Two tiers, run separately | 0% and 6% | $4,423.53 | – |
| One average, rate weighted by value | 3.6% | $4,272.86 | $150.67 |
| One average, rate weighted by bottle count | 1.2% | $3,380.08 | $1,043.45 |
Averaging by bottle count is the natural mistake, since most of the bottles do not grow, and it understates the result by $1,043.45. Weighting by value is much closer but still short by $150.67, and it always will be: compounding rewards the fastest-growing part of a collection more than an average rate can capture, so a blended rate understates the spread every time.
Why it matters even if you never sell
- Insurance. If you insure a cellar on a projected value, the error lands in the bottles you would most want covered.
- Deciding what to drink. Separating the tiers shows you which bottles carry the value, which is useful when choosing what to open.
- Seeing the assumption. Two runs make it obvious that nearly all of the projected gain comes from 12 bottles and one guessed rate. One blended run hides that.
How to do it
Split the cellar into groups that you would expect to behave alike: wine to drink in the next year or two, wine to age for drinking, and any bottles you think might be worth more later. Run the estimator once per group with its own average and rate, and add the results. Use a rate of zero for anything you bought to drink; for everything else, try several rates, because the honest answer is a range, not a number. And remember the costs that sit outside any of these sums, covered in the guide to what compounding leaves out.
Work it for your own bottle
The figures above come from the same code the calculators run, so typing these inputs in reproduces them exactly. Your own bottles are the point, though: run each tier through the collection value estimator.
Not investment advice. This is arithmetic on assumptions you choose. Wine is an illiquid, unregulated market, past prices say nothing reliable about future ones, and nothing here is a recommendation to buy, hold or sell anything. Speak to a regulated adviser before treating wine as an investment.