The correlation between value and any network statistic (difficulty, blocktime, hashrate) is poor at best. Getting that info from a source outside the network (exchange etc.) is pretty sketchy. How do you propose to make the loyalty mechanism at all sensitive to value?
This is an excellent question. I do not think we as a cryptocurrency ecosystem have a stable universal API yet to pull coin exchange information to build into a coin as yet - it would be unwise to build a coin on the API of specific exchanges - so we have to rely on other factors.
There are factors which we can obtain with 100% certainty from the blockchain itself - which avoids reliance on any external data - which can be helpful. They are:
1. Current difficulty level - which can be used as a reasonable stand-in for the aggregate network hash rate at the start of the current difficulty level
2. Previous difficulty level(s) - which can be used as a reasonable stand-in for the aggregate network hash rates at the start of respective difficulty levels
3. Which wallets address received block awards during the each of the current and previous difficulty eras
4. Anticipated forthcoming difficulty level if the change were to happen now - which can be used as a reasonable stand-in for knowing the aggregate network hashrate at present
Although we cannot derive current market value based on these factors - we can use these indicators to pinpoint when there is or is about to have high demand to mine the coins - which is a pretty reliable indicator that the market value is much higher than it was in the immediate past relative to difficulty level. It's kind of like being a bat and flying on ultrasonic sound reflections. We can't directly see the objects, but we can sense the feedback, which is enough to avoid running into them.
Let's say we are at difficulty level 90, and we have a total of 1 GH of mining going on (I am sure the correlation is not correct, this is just to illustrate the idea), and based on this hashrate, the network anticipates adjusting the difficulty level down to 25 (factor #4 above). We also know that the network difficulty level adjustment does not take into account that people will tend to want to mine more when the difficulty goes down, and will want to mine less when difficulty goes up. In other words, when the difficulty goes down to 25, if the hashrate stayed at 1 GH, then the block will be solved every 10 minutes as desired. But the fact the difficulty is at 25 will inevitably cause the hashrate to go up - we have never seen it not do this.
We can also reasonably surmise that the 1 GH of hashing, will continue to hash once the difficulty level drops to 25. But there will be additional hashing power that will show up in response to the easier difficulty level.
By using wallet address information (factor #3 above), we can programmatically construct a list of miners (represented as wallet addresses) are members of the stable 1 GH of hashing power, that was actively hashing when the difficulty was at 90 (and we can look back many adjustment eras, to include more miners, perhaps solo miners that don't find blocks very often, in this set). So the entire network can be looking at the same list, 100% derived from analyzing the blockchain.
During this difficulty level 25 period, we know difficulty has dropped more than 50%, so we know we want to limit rewards for cherry pickers. During this exceptional era, when a new block solution is submitted into the network, the network can compare the wallet address of the node that has submitted the solution, with the list of wallets programmatically constructed above which acts as a kind of loyal miner white list. If the solution submitter is not on the list, the entire network can agree and have consensus based on a preprogrammed formula to reduce the block reward given to that wallet (such reduce by 50%, from 50 coins, to 25 coins). If the wallet submitting the solution is on the list, then we can agree as a network, that we award the full 50 coin award. This is oversimplifying a bit - in reality we'd want to make it so if someone solved a block once, they can redeem during the easy period once, if they solved 10 blocks - then they can redeem 10 times at full value, etc. But you can see the principle.
What this formula would accomplish is not directly to see price swings - but like the bat - it can sense ultrasonically indicators of coin hoppers coming to collect easy coins - and with 100% precision, snipe half of the reward away from just the cherry pickers - while allowing proven members of the set of miners who have been contributing hashes (by solving blocks) during more difficult times - the benefit of the full award. Now, my original formula was to reduce the reward by a fixed 50%, but if the coin price jumped up 100-fold, it would still even with the 50% reduction, attract a lot of coin hoppers. So the formula could be adjusted, so it takes the ratio of difficulty levels and uses that. So in our example, if our difficulty went from 90 to 25, we meet the criteria of an exceptional reduction in difficulty (it went down >50%), which engages the logic. We would mathematically derive that 90 => 25 is a 72.7% reduction in difficulty, which the whole network can agree to. So we would grant a full 50-coin award for solving blocks during this difficulty era to any qualifying wallets, but reduce the reward by 72.7% for any other wallets that solve the block (so the award would be 13.65 in this case). This makes the newcomer see the exact same award as if they had showed up during the higher difficulty era, while those who had been hashing gets to see the easier difficulty era. This entirely defeats the logic of showing up when difficulty gets easier, to mine for a short while, then go mine somewhere else. Now, if the coin valuation went way up, say 100x, then the newcomer would still come and mine the coin anyway despite the reduction in award, because it is still profitable (13.65 coins that are 50x more valuable, is still worth 682.5 coins at the previous market value). If the network hashrate swelled to 10x the previous levels based on this difficulty (9 new mining GH for the 1 existing GH), it means the block reward would look like this - 36 blocks x 10% = 3.6 blocks receive 50 coins each = 180 coins to the loyal miner. 36 blocks x 90% = 32.4 blocks receive 13.65 coins each = 442.26 coins to the newly arrived miners. So a total of only 622.26 coins got produced during this exceptionally easy difficulty cycle - limiting inflation. This is much better than 180 coins to the loyal miner to 1,620 coins to the newly arrived miners (total 1800 coins produced = lots of inflation) which would otherwise prevail. Since more miners showed up despite reduced payment we can safely assume it is in response to rise in catcoin value in the market. The difficulty then adjusts upward, and everyone can start receiving the full 50-coin awards. But if this was a case where it was only a temporary reduction in difficulty, and it was going to go up again - or there's going to be a longer-term reduction in difficulty level - then the newly arrived miners presumably leave and the stable 1 GH of miners stay, and we have saved ourselves 1,177.74 coins of inflation and coins being dumped in the market, during that difficulty era, and we can keep the same restriction in place for as long as the difficulty stays in the same ballpark. Realistically, people respond to incentives and the 1 GH hashrate would stay around there inherently despite the reduction in difficulty from 90 to 25 UNLESS the value of the coin went up in the market - in which case people would be attracted to mining the coin for its market merits rather than immediate dumping profits -and would stay and continue mining at the next significant difficulty increase - at which point they have earned the right to get full credit for solutions submitted from that point forward - and we can count them as a genuine increase in the baseline mining GH of the coin deserving full reward, and the difficulty increase is a genuine increase, not a temporary spike. So by this means, network hash rate will increase in response to increase in the coin value in the markets as we want - but not in response to just the difficulty level dropping while the coin value remains the same or dropping (which is inherently a temporary difficulty drop anyhow). By this means, we eliminate the tendency for coin hoppers to create and sustain wild oscillations in the difficulty - while still allowing the coin to respond quickly to rise in market value of the coin, by allowing new legitimate hashing power to be added to the coin.
I hope that presentation was clear enough that you could follow it, I know it's not exactly white paper quality - I wanted to get you a timely answer. Please feel free to ask me to clarify any aspect if I didn't do a good enough job.
Etblvu1
, you will need to contact the exchanges with your updated wallet. 
