I want to raise a particular concern that I made in a comment earlier, and I would like to see it refuted so that I would be more on-board with the reverse conversion plan that's being floated around.

Specifically, this comment, thanks to @twodollars for starting up the thought exercise.

The Pump and Convert Scheme

Initially, @twodollars came up with a very simple mechanism, only involving a pump:

Let's say P(t) is the STEEM price feed over time.

1. Buy/Have X amount of STEEM.
2. Start 3.5 day [reverse] conversion at 1 STEEM --> P USD
3. Buy more STEEM to pump price.
4. Now have P(t_4) * X SBDs.

Now imagine that we are already at a stable peg situation where SBD is worth 1 USD, and let's assume for argument's sake that this peg is maintained throughout the course of this scheme.

Let's say you started with Y dollars, and allocate Z dollars for pumping. In that case, you'd get X = Y/P(t_1) amount of STEEM, and at step 3 you would have greater than Z/P(t_4) amount of STEEM. At the end of this procedure, you now have

P(t_4) / P(t_1) * Y SBD   and >= Z / P(t_4) STEEM

This means that depending on how successful your pump is, you've already multiplied your value in SBD holdings, which you can cash out on exchanges assuming the peg holds. It looks like in this scheme it is uncertain how much you'd be able to recover of your STEEM holdings, but if your values are chosen right, perhaps that value is not significant (if say the amount necessary to sustain a pump for 3.5 days is much less than the initial input).

Feel free to read the original thread to see the original description of the scheme, which I am attempting to lay out. It gives sample numbers as well.

My first reaction to this was that, if this could be carried out, why couldn't it be carried out in the forward conversion? It's because the two conversions are not symmetric at all. It's based on the price of STEEM, and if you tried to do this the other way, it requires you to tank the value of STEEM, and you need access to STEEM on the market to do so.

Going Both Ways

I then thought about how to make this even more fun, and came up with the following:

Let's say P(t) is the STEEM price feed over time.

1. Buy/Have X amount of STEEM.
2. Start 3.5 day [reverse] conversion at 1 STEEM --> P USD
3. Buy more STEEM to pump price.
4. Now have P(t_4) * X SBDs.
5. Start 3.5 day conversion back. 1 SBD --> 1/P STEEM
6. Dump the STEEM to tank the price.
7. Now have P(t_4) / P(t_7) * X STEEM.

Depending on various factors, you could potentially turn a profit even from steps (3) and (6) alone, though I don't really know, because that's no different than a traditional pump and dump. Can it work for a currency with the current maturity of STEEM? I do not know. But you can see that after this process, even a slight pump and dump could boost your STEEM holdings by a significant margin, especially if you keep repeating the process.

Note that if this process is doable, adding a spread/fee does not help, unless it is a significant penalty that the likely pump and dumper has to evaluate the risk of achieving its price targets vs the penalty.

Remember also that one of the caveats here is that for the price feed, it takes a 3.5 day average of the prices, so there is a bit of lag. In theory this procedure could still be carried out, you would just have to sustain a pump for 3.5 days and a subsequent dump for 3.5 days. The other mystery factor is that witnesses control the price feed, so if they wanted to take action against this pump and dumper, they could. That threat could also be a deterrent, though I would discount that because it requires the witnesses act in lockstep and react immediately to this event.

The final risk is that other market movers may have other ideas. You need to be confident that you have enough to move the entire market, and the higher the market cap of STEEM, and the more spread the distribution of STEEM is, the harder that becomes.

Are my stated caveats enough of a deterrent?

Addendum

This post has been edited after I understood @twodollars setup more, and I have added his proposed scheme in the first section.

From what I gathered / read, pumping the price up is a relatively easy task, and the hard part is the dump without losing money. So in fact, in my portion of the scheme, it's likely that between steps (3) and (6) the schemer may lose a significant amount of money. But it may very well be offset by the amount of gains in the STEEM token achieved. If this carries out during a time the peg is actively maintained, then you could cash out of this procedure through SBD, as mentioned in the first section.

---