There's a hell of a lot of reasons people are selling.
The periodic sells of 1000+ DRKs that were ticking like clockwork for days, were not "people". More like a single whale exiting. Actual people (meaning a diversity of investors) weren't even selling even while the price was being driven up to 0.025 and liquidity was quite dry.
ROI is heading straight down and you can expect it to continue along that path. Sure 18% annually sounds OK, but the price would have to stay consistent for that year which seems to be a bit of a stretch considering drk has played all its relevant cards and the future of btc is looking a bit grim.
There is nothing grim about the future of BTC. The only "grim" thing are the multitudes of trolls infesting bitcointalk to create this specific sensation.
Less than $2 a day and dropping fast is not really that enticing for every $3,500 you invest in such a risky investment like an altcoin.
It depends how you calculate risk. If you see a few million marketcap as a ceiling when real world adoption takes place, ok. If you believe there will never be real world adoption either, ok. After all, BTC was trading at single digits for years because people couldn't see it in terms of real world adoption (which is starting to happen).
It has functional privacy, but thats really only relevant in the very short term because of the whole trusted third party aspect. I ignored it for a long time and I believe you can too, as long as you keep focusing on the pipe dream.
It would be trusted if one only did one round of mixing. That's why you go over several rounds of mixing were probabilities go down.
Every single anonymous solution makes some trust assumptions. Cryptonote doesn't have a third party performing the mixing, yet if you go over at Monero's thread you'll see how they attack Bytecoin for having a premine which deanonymizes future anonymous transactions. Large control of outputs = deanonymization. Is there a "trusted third party" there? No. But all mixing systems depend on having legitimate mixing partners. So how will that work?
Even when using zero knowledge, you have to trust that someone doesn't have -or won't have in the future- a quantum computer, or that they have really thrown the key away so as not to generate infinite coins on demand.
And even Bitcoin's trustless nature, which isn't anonymous, also makes certain trust assumptions when people are throwing their hash into centralized mining pools that could easily go over 51% and manipulate the trustless nature of transaction handling.
Evan is a loveable genius, but the coin is completely reliant on him which causes centralization. THis wouldn't be a big deal but it means he is vulnerable to being influenced and making reckless decisions.
This is an area that can be "attacked" both ways.
If Evan stops => "Oh DRK is dead"
If Evan continues => "Oh DRK is centralized on Evan"
How can it be that a half-full (or half-empty) glass, always appears empty? There is a problem of perspective here.
And where does drk go from here? Masternode blinding is nice, but is still relying on third parties.
Rounds of mixing + blinding are there to counter the trust issue into a probability that is not even ...real.
Instantx is cool but its really just forging confirmations, and you can tell how much the crypto community trusts th is apporach by the fact that zero exchanges have implemented it. COncerning the scalabitlity issue... it is solving a problem that does not exist for drk.
Just because something isn't experiencing widespread adoption immediately, doesn't mean it's useless. These things occur gradually.
InstantX is not just fast confirmations. It's a tool that can cut the middle-man from BTC transactions. With some bitcoin payment processors you can buy something instantly with zero confirmations. But the merchant relies on the payment processing company to take the risk for him. With InstantX the merchant doesn't even need the payment processing company and their zero-confirmation mediation. That's a nice thing to have in crypto, and even keep payment processors honest from requiring a few % per transaction in the future.