I may not know the technical feasibility of this. Also, I don't know whether this idea has been put forward before, but here it is:
Create a cryptocurrency whose block reward, difficulty, target block generation rate, and transaction confirmation time change* as an average of all known cryptocurrencies.**
Let's call this hypothetical cryptocurrency ArbitrageCoin (ARB).
How it might do this is to connect to all (or carefully selected) cryptocurrency peer-to-peer networks on a "read-only" basis, and learn the generation critera of each currency, and then dynamically alter its own criteria as an average of every currency it finds.
Very simple example: say that ARB connects to cryptocurrency networks CCA (CryptoCurrency A), CCB, and CCC. It learns:
CCA current block reward 25, difficulty 4M.
CCB current block reward 50, difficulty 2M.
CCC current block reward 100, difficulty 1M.
It then averages these critera --
Block rewards (25 + 50 + 100) divided by number of criteria to average (3) = ((25 + 50 + 100) / 3) = 175 / 3 = 58; ARB thereby knows that the average block reward of all cryptocurrencies (which it has scanned) is 58, and ARB sets its own block reward *to* that average; ARB's block reward thereby becomes the average of CCA, CCB, and CCC, which is 58.
It does the same thing for the difficulty levels it found on the network:
((CCA diff. 4M) + (CCB diff. 2M) + (CCC diff. 1M)) / number of criteria to average (3) = diff. 2.33M); ARB thereby sets its own difficulty level to 2.33M, or the average of the difficulty level of the three cryptocurrencies it scanned.
What could be unique and useful about this?
Assuming it does this for all cryptocurrencies which are deemed "mature" in any way--however that would be measured (which would need to follow strict and useful criteria), if ARBs generation criteria are an *average* of all those cryptocurrencies, any of ARB's generation criteria can potentially go down at any point in time when some cryptocurrency which it scans becomes a statistical outlier.
Simple example scenario; difficulty average ((4+4+2+1)/4) = 2.75 changes to ((4+4+2+1+1)/5) = 2.4, or -.35 of what it was. How might that happen? If a newly "mature" cryptocurrency, CCD, joins ARB's multi-peer-to-peer network, by virtue of CCD fulfilling maturation criteria.
Another scenario: if the actual block generation time of any cryptocurrency (which ARB monitors) goes sharply down, while the actual block generation time of three other cryptocurrencies goes sharply up (they are mined faster), the average block generation time will go down, and because ARB strictly follows that average, ARB's difficulty will adjust down to that average, to re-target for faster block generation.
How might that be useful?
Because ARB would be generated faster, it would be possible to mine more ARB in the same time frame, and thereby sell them collectively for a higher value (until cryptocurrency exchanges catch on to the disruption, and correct in response), which higher value would have the effect of disrupting other currencies downward (say, CCD, for our example), because here suddenly another currency (ARB) has more relative purchasing power . . . until CCD corrects upward. Why would CCD correct upward? Because when CCD's value (price) went down, it became more of a "value buy" for people in the market who expect to buy CCD it while CCD's price is lower, and then sell CCD when its price is higher.
Also, there could be a variety of other uses for a crytpocurrency which averages against others. If ARB, because it is averaged, has relatively slower price change fluctuations than other, more highly reactive currencies, ARB could become a medium for arbitrage.
Example: Cryptocurrency exchange X sells currency CCA for 50 ARB, while exchange Y sells CCA for 60 ARB. Users could transfer ARB to exchange X, buy *more* units of CCA (because it costs less), then transfer those same units of CCA to exchange Y, and sell them for 10 more ARB. This scenario may be more likely to occur with a currency that is relatively stable or slower-moving versus other currencies; if ARB is dynamically averaged, it could become such an arbitrage medium (at times).
(This would also help solve the problem of trying to use U.S. Dollars as an arbitrage medium; common arbitrage strategies follow the pattern just outlined, except that they use U.S. Dollars instead of ARB. But Dollars can be much more slow and difficult to move between exchanges than cryptocurrency (like our hypothetical ARB).
But at the same time this could generate arbitrage opportunities, it can make those arbitrage opportunities a bit more risky (and therefore potentially more profitable to people who exploit that risk the proper way) . . . because ARB is subject to correct as an average to multi-currency (wider) market disruptions; in other words, if a lot of currencies "correct" in a short enough time frame, they would collectively more swiftly move ARB (their average) toward that correction.
In other words, ARB, if it is created and adopted widely, may probably cause random, moderate inflation (loss of value) among cryptocurrencies--which would actually be good for cryptocurrency, if healthy inflation is followed by healthy value gains (which I believe it is), AND/OR it may probably cause random, moderate appreciation. Or it could swiftly cause either.
I can't code this. If others think this is a good idea, I welcome them to create it and try it out.
Thoughts? Disagreements? Education I need?
Extrapolations? Further ideas? I welcome these . . .
*What other criteria could change? And which of these can even be learned directly from a peer-to-peer network--or are there other reliable ways to learn these things?
**Or perhaps of all known cryptocurrencies which meet certain criteria, for example any currency which may be termed "maturing" in any sense, by measure of how many generated blocks, nodes, and transactions are (or have been) on the network for that currency.
