The investment-grade exceptions
A small set of references have produced genuine investment-grade returns since the mid-2010s. Rolex Daytona 116500LN: 2x retail in 5 years. Patek Nautilus 5711/1A: 4-5x retail at peak (2021). AP Royal Oak 'Jumbo' 16202: 1.5-2x retail. Lange 1 Datograph: 30-50% gain in 3 years. These references combine deliberate supply constraint, cultural cachet, and the post-2017 collector boom that drove the watch-investment thesis.
Why most watches don't compound
Most luxury watches lose 30-50% of retail value in the first 2-3 years of ownership and continue declining slowly. The exceptions list above is small; for every Daytona that doubles, there are ten references (TAG Carrera, IWC Pilot, Omega Speedmaster Reduced, Hublot Big Bang non-LE) that depreciate predictably. The marketing-led 'watches as investment' narrative selectively highlights the winners; the median watch is a depreciating asset.
The hidden costs
Service every 5-10 years: CHF 600-2,500 per service, depending on tier and complications. Insurance: 0.5-1.5% of value annually for adequate coverage. Storage: safe deposit box or home safe; modest cost but a consideration for collections. Liquidity discount: selling fast (days) means accepting 10-20% below market; selling at full market price can take months. Net: holding a watch for 10 years carries 10-20% in genuine costs that erode investment returns.
What to actually do
Buy what you want to wear. If the reference also happens to be supply-constrained and likely to hold value, treat that as upside rather than a reason to buy. Don't speculate on grey-market premiums; the 2021-2022 watch market correction wiped 30-50% off the most-speculated references in 6 months. If you must invest in watches, focus on the 5-10 references with provable 10-year track records (Submariner, Daytona, Nautilus, Royal Oak, Lange 1, certain Patek Calatrava and Vacheron Patrimony refs) and accept that you'll wait years for liquidity.