What I think about this is that there is a certain market for bitcoin. Namely for all those interested in privacy. For whatever reason - they may be amoral, simply illegal, half-legal, but that does not decrease their role in the global markets. This role will increase with the government fighting against cash to increase their control over the sheeple. Given that among those who want privacy for their money are also US secret services, and the US itself too (as long as the dollar is stronger than the average currency, bitcoin helps the local elites to transfer money out of their countries mainly into the US), there will be some long enough period where the US, as well as the EU, will not openly fight against the bitcoin. During this time, it will be increasingly used by some big players - lobbies, capital transfer of big firms, bribes to politicians everywhere, the mafia, the CIA.
Let's note that there are some weak points of bitcoin which make them unusable for the everyday payments: There are actually much too large costs per transfer, and a long time for confirmation so that they are unusable for buying a cup of coffee or so. Not sure if this can be solved, the large payments per transfer can be reduced, I think, they are probably simply a consequence of the heavy increase in value, and the fact that they are in bitcoin units so that they have increased together with the bitcoin. But, even if they remain, they are irrelevant for big players.
So, there is a remaining window of opportunity for bitcoin (until the whole world unifies to fight bitcoin) and a large enough market share (all the criminal payments) which gives a large enough amount of turnover of bitcoin. It is this turnover which defines the value of the bitcoin. Given that actually you need a day for a safely transferred bitcoin, the turnover speed will remain bounded. You will not make a speculative trade with buying and selling seconds later in bitcoin. Usually one can expect that if you want to buy something, you will hold the money for this in the relevant cash around a month or so. If some illegal market has started to use bitcoin, there is a high probability that they will not stop doing this, at least not before all the states of the world unite to fight bitcoins. And this last thing becomes not very probable if the market for bribes to politicians has switched to bitcoin. This is a sufficiently large market, it secures the existence of bitcoin in the same way as it has politically secured in the past (and secures even now) all the "tax havens". I bet that the big players in the drug business were also clever enough to switch to bitcoin for international money transfers too.
It is, of course, very difficult to make expectations about the turnover in such criminal activities. Whatever one thinks about this, the most reasonable way to make expectations would be to make estimates of the turnover per person. One can expect that for every illegal activity paid actually in cash there will be, behind, somewhere a corresponding bitcoin transfer of the same size. So, make expectations about how many drugs are sold by the average person. Whatever the result of this evaluation, what follows is mathematics. Say, you estimate X \$ criminal turnover per person -> X \$ bitcoin turnover per month, X \$ hold by those who make these payments and compare this with the 25 Mio bitcoin / 8 billion people = 0.003 bitcoin per person. This gives the rate of the bitcoin of X \$ = 0.003 bitcoin. So, the actual rate, say, 15000 \$ per bitcoin, corresponds to X = 45 \$.
Beyond the illegal markets, there are other fields attractive to bitcoin. First of all, savings. I have heard that people in Zimbabwe have started to make their savings in bitcoin. Natural, to make savings in the local currency would be stupid, savings in cash at home are vulnerable to theft, but savings in bitcoin can be protected against theft. Now, that means if the average person decides to use bitcoin for 45 \$ of his savings, the actual rate is not speculative but safe. And there are other (legal or half-legal) markets, namely all international payments, where bitcoins may be used.
Note that the turnover of bitcoins is something which is open information (you can extract this information from the blockchain data, but I have to admit that I have not done this, too lazy), while the amount of savings is harder to estimate. But part of this is also accessible via the blockchain, by observation of the amount of bitcoin which are frozen in accounts which are not used. This combines accounts used for savings with accounts of bitcoins simply lost for technical reasons. But this does not matter much. If they are lost, they reduce the available bitcoin, thus, reduce the 0.003 bitcoin per person, thus, increase the value of a bitcoin permanently. If saved, they could be broken into the market in a bitcoin panic. So, it would be even better if they would be lost, and counting them as being saved is a safe bet. AFAIK, bitcoins not used after 2011 are around 3 Mio.
I have a quite funny evidence that politicians are already using bitcoin for illegal activities. Namely, Ukraine. There have been laws fighting corruption (enforced by the West - they have recognized that Ukraine with the actual level of corruption is useless for them, because they cannot support the government in any way, any support given to the "government" is simply distributed among the politicians for nothing in return), and these laws obliged the parliamentarians to declare all their income and property. Of course, almost all of them have, now, quite a lot of property and no plausible legal way how to get it. Of course, most of what they really own they will simply not declare. But, of course, if you have a possibility to declare some part of it (at least the visible part, namely the big car they want to drive to show they are big guys), one will use it. So guess how many of the Ukrainian parliamentarians have been able to buy such big cars? Correct, they have (or at least declared to have) successfully speculated in bitcoins.
I would, therefore, recommend you to make your own expectations about the amount of the regular turnover of illegal and half-legal markets made in bitcoin, as will private savings in bitcoin, all this in dollar per person, and then to compute the rate by comparison with the 0.003 bitcoin per person.
Do this in three variants - first, an expectation for the actual turnover/savings, to compare it with the actual rate, to estimate how much of this is speculation and how much real. Second, the variant that bitcoin remains/becomes the established leader for these activities in the future, to estimate the rate which will be maximal one without speculative bubble effects. Third, the variant that the states nonetheless unify to fight the bitcoin. (But there is no real need to estimate this, you can safely put a zero here, and consider everything you have in bitcoin as speculative with the danger of full loss.)
About the speculative part: The diagrams for the bitcoin value can be, approximately, seen as a repetition of the following scheme: Add a slow permanent (exponential, but with a fixed rate) and the usual pattern of a speculative burst (exponential growth with a much higher rate, followed by a bust which reduces it to zero). Actually, we seem to be at a zero after one such speculative bust. The picture above shows these bust peaks in June, Sept., Nov., Dec. But this scheme I have seen and described in these terms already before this. My interpretation is that the slow permanent growth describes the increase in the real value based on turnover and savings made in bitcoin. If one follows this, there would be no reason to sell bitcoin if they are down after some burst. In particular, now is not a time to sell.
The most relevant of these considerations I have made, hm, some time ago. I'm quite satisfied with the results,

as you may guess, but to make the resulting decisions was, on the other hand, much easier than now. Except for the guess that bitcoin will win the game, which is more plausible today.
Based on my own estimates, I have sold yesterday 0.1 bitcoin, but this is not in contradiction with the point above that it is not the time to sell now, but simply profit taking combined with the idea of diversification: If you have a speculative investment, take some part of the profit if it raises, and put it into safe things like gold. Once you take only a part of the profit out, the part which remains there increases in value anyway, only slower than the speculation itself. That means, you continue to profit from future raise, but, on the other hand, you will lose less if it all goes down, and your behavior decreases the bubble in the best possible for the society way - slowly, decreasing the overheated market. Note that if you decide to participate in a speculation, you also have to develop a strategy how to realize your profits. Else, you will not realize them, but lose everything once the speculative market crashes.
As a consequence of following my own profit-taking strategy as described above, I have now already what I need to remain an independent scientist forever, even if everything I have in bitcoins now becomes zero in one day.

Anyway, I have some bitcoins yet, so that I'm interested that bitcoin raises in future even more, so, evaluate the considerations above having in mind this interest. I would start selling the remaining bitcoins to minimize losses if the bitcoin goes below 9000 Euro, and sell them all at 6000 Euro because if such values would be reached, it would be clear that there is something wrong with the considerations above.