The vintage cymbals I was speaking to are the Zildjian Ks. I originally had that written but changed it to keep this a bit more general ... as I also believe it applies to classic Paiste. The interesting point here is we will see an increase in Sabian handhammered value as these others decline. It will be an interesting thing to watch over the next two years.
I'm glad to see your qualification of your original statement. I'd say the used cymbal market is not a single monolithic market which moves in unison. My statement is evidence based, and I have the numbers. I've been recording all eBay sales of 602s and Sound Creations since December 2005. I've just begun a new round of analysis based on Dec 2005 to end of October 2011. I don't record Istanbul Ks (that's Bills territory) and I haven't really been recording early Sabian HHs completely enough to see what's what there (although I agree with you that they haven't been "discovered" for as long as Istanbul Ks thus they are likely to have some upside on price).
Analysis of the first 3 years Paiste data
http://www.cymbalholic.com/forums/showthread.php?30271-New-Paiste-602-SC-value-rule-of-thumb
demonstrated that there are significant differences between the price patterns for different models and sizes of 602s and Sound Creations. Some go up, some go down, some stay the same over the same time period. The only cymbal among all the models of 602s and Sound Creations which might be a "good investment" is the 22" 602 (transitional) Dark Ride. That is no doubt rarity: demand still exceeds supply. The 22" Sound Creation Dark Rides might appear to be a "good investment" but not when you take inflation into account. Istanbul Ks remain the only "investment" I've seen which have done better than inflation (based on new prices in the 50s and 60s versus what they fetch today).
However, I'm just starting the analysis phase for the full 6 years of data and it will take some time before all the results are out. The market is extremely complex, and is driven by a number of factors. The statistical modelling is also very complex. And even 6 years of data isn't enough to predict what will happen in 10 years time (which might be your investment horizon). But I've been lucky in that I've been monitoring over two (or some might say 1 double dip) recessions, and also during the reintroduction of some models and sizes of the 602 series. I've seen just a few used sales of reissue 602s, so I'm waiting for more evidence to see how if any the reissue might impact the used market. I've seen different opinions, but not enough data to test them with so far.
Now "all eBay sales" isn't the same as the complete used cymbal market, but it is the best proxy if you are a retired statistician and have no research grant to undertake a better research design. I began recording this data partly out of interest in cymbals, and partly to build up a sufficiently rich and freely available data set so I could experiment with different modelling techniques and share all the data with others. Over the years I've worked for big multinational corporations and governments, but all that data is proprietary and not allowed in the public domain. This data is free and open source. The spreadsheet of raw data is available to anybody who wants it. Just ask.
Here is a pdf of where I'm up to on the very first leg of the reanalysis
http://black.net.nz/cym2011/pcalc2.pdf
which has only come as far as rides as of last night. You can see all the rows I have left to fill in (the blocks of zero below the rides are to make sure I don't mess up when I copy and paste chunks). This is only a preliminary pass, and is part of the data cleaning process (there is always dirty data). I also usually prefer to use medians since they are less susceptible to outliers. But if you are a rocket scientist, you knew that already. Yes Sir
Another goal of this initial analysis is to get a feel for how much things might have changed during 2010 and 2011. It would actually be great to decide that things aren't changing much over time, because then you can pool all the data (as in the first columns of figures) and get decent sample sizes for rare models and sizes. Longer term analysis will look at trends to see if GFC mark 1 and mark 2 show up in prices. But you have to understand all the other factors which affect the market and control for those before you look for a GFC effect. That's the trouble with observational vs experimental science. You have to crank up the serious statistical work. After you've done all that and controlled for every other factor (whew) you could start to look at the separate effect of "market maturation". Am I right that the effect which you are proposing is some "market maturation" model RogerSling? You may have a better term for it than I've thought of off the top of my head.
Oh, and I forgot to mention that the Spizz (and individual cymbalsmiths) are another market which needs to be separated out from the big 3, and the gaggle of Turkey based suppliers, and those Chinese. Again I don't religiously collect data on these (life is too short) but Spizz has caught my attention because I seem to see used ones selling for more than new ones even though they are still being produced by the Don Roberto. That's unusual, but it is confounded with massive exchange rate changes (among other things).
