Topic: [BTC-TC] Deprived Mining Speculation (DMS) - page 59. (Read 198958 times)

If you need someone for testing purposes let me know.

I'm working on a library to access the BTC-TC API, which I will make available as opensource. I've got most of the features already implemented. There will be some programming legwork required to actually turn it into a useful trading client, but the library should handle the annoying details of talking to BTC-TC.

Is there an open source trading client for BTC-TC?

The only ones I've seen running so far on BTC-TC/LTC-Global are ones that just trade the spread with tiny quantities.  Some of those are exploitable when the spread gets wide enough (which doesn't happen on DMS - you generally need a 40%+ spread because of the parameters they seem to run with).

Random work? That sounds like my kind of work ^^

We can use the random walk algorithm.

The actual coding of a bot that performs automated investments/trades is not the difficult part. The real challenge lies in determining an algorithm that yields profitable returns and limits potential losses, even if the market makes unexpected movements.

Interesting bot! Now you need to code one that invests and trades for me, hands free.

It was written specifically for this.  It's in C#. Here's a screen-shot of it :



Grayed out section to the right is the relevant part - allows me  to run in test mode (where it reports what it would have done but doesn't do it), only process trades for 1 user (used in testing), set the frequency it runs at, set it to only do automatic transfers below a certain amount and set a start time for it to catch up on back-log from (by default it only process trades that occur after it's started).

I sent in a few of my own PURCHASE just before the screen-shot so it showed what it looked like when a transfer actually occurred (last real one was some hours back).

Which bot do you use? Can you give some details or a link?

Auto-transfers are now fully enabled - though being watched carefully by me.  So if you send PURCHASE and don't receive your MINING+SELLING within 5 minutes drop me a PM (though check BTC-TC hasn't gone down first).

If BTC-TC does become inaccessible (which happens occasionally - it was unavailable for me for a while yesterday night) then the bot should catch up with any back-log as soon as connectivity returns.  This net connection is also not the most steady and, for various reasons, it isn't set to reconnect when connection drops.

From time to time it'll be turned off as it's running on the computer I'm using (so I can monitor it) which tends to get restarted a few times each day.  Once it's been goign a few days without problems I'll sling it on a new computer with its own internet connection so it doesn't have problems whenever I max out bandwidth usage on this connection myself (I have a backup mobile broadband connection that's currently unused which it can have for itself).  At that stage it'll be set to reconnect automatically and downtime other than when BTC-TC is down should be far less frequent and for shorter periods.

It may also be down occasionally whilst I test software upgrades (LTC-Global integration is being worked on to allow cross-exchange transfers that are needed for the next securities LTC-ATF will be launching).  Functionality is now added (and tested) so that after any period offline I can set it to clear back-logs.

One last point is that it does NOT attempt to retry transfers that fail (just reports them for me to manually fix).  That has to be the case as I can't rely on everyone being honest and returning double-sends (on occasion BBTC-TC has been known to report a failure when an action actually gos through - e.g. yesterday it 3 times failed on transferring my DMS management fee with ''wallet lock" errors but two of them actually transferred.)

My modelling showed that most of the difference to the value of DMS.MINING occurred in the first 3-4 DIs of the projection, i.e., a few big DIs now followed by smaller ones later on affected profitability much more than the reverse. MINING is a very risky thing to hold. It was easy to see when DMS.SELLING was cheap (i.e., when it was fair value assuming 8% DIs, but actual increases were 19-20%); I don't think we'll ever be able to say the same thing about DMS.MINING.

Auto-transfers back off whilst I'm out for a bit.  So any transfers now won't get processed for a few hours.

Auto transfers will be turned on for next half hour or so - but only for quantities of 25 or less.

If anyone buys any PURCHASE in that period and wants to help test, then please split them up over a few smaller transactions and we'll see how the bot handles it.  Will post again once its turned back off (I'm heading out in half an hour and not gonna risk leaving it unattended yet) so after that DON'T split transfers up - as I'll have to catch up manually (bot has no current functionality to clear a back-log).

It's definitely true that the profit potential on SELLING has dropped.  But IF the price is correct then it's not as bad as it may at first seem - as for its value to really be that high the bulk of that capital would need to be returned pretty promptly anyway.  Part of the design of DMS was to try to ensure that even in this scenario (anticipated very high difficulty rises, so low MINING price and high SELLING price) SELLING didn't have to leave capital tied up excessively.

But if the prices are precisely correct then NEITHER is worth holding.  If total information were available and time-cost of capital and CP risk ignored then MINING would trade at slightly under total life-time expectation of dividends and SELLIng slightly above (reflecting that SELLING has the EV from investment).  The margin of profit on both would be so tiny as to make them fairly worthless as investments due to any cost placed on CP risk making them -EV.

DMS relies on the situation not arising where everyone agrees on exactly what the prices should be - but if that ever happens then I'd expect a mass sell-back as DMS would no longer be serving any purpose.

Regardless of how close MINING is to its true value, what is already true is that SELLING is reaching the end of its profitability. SELLING is trading around 0.023, MINING at 0.005. Which means that there is at most a 20% (and a little bit) profit to be had for SELLING (plus any gains from investments and PURCHASE sales). Compared to the 0.033 that SELLING has earned so far in dividends, that's not that much.

In not too long, the MINING/SELLING interplay will become quite a bit more delicate, instead of the "Buy SELLING and profit" motto that was true in the past (and might still be for some time to come).

Well it just did its first actual transfer (I've got it running in a mode where it only actually does transfers to my personal account - everything else it still just reports but doesn't act).

From its log:

Most recent entry in trade history at 08/08/2013 18:05:50 (server time)
Update started at 08/08/2013 18:07:57
TX-IN: 20 x DMS.PURCHASE <- DeprivedPersonal
TX-OUT: 20 x DMS.MINING -> DeprivedPersonal
TX-OUT: 20 x DMS.SELLING -> DeprivedPersonal
TX-IN: 30 x DMS.PURCHASE <- DeprivedPersonal
TX-OUT: 30 x DMS.MINING -> DeprivedPersonal
TX-OUT: 30 x DMS.SELLING -> DeprivedPersonal
Update completed at 08/08/2013 18:08:06
Most recent trade was at 08/08/2013 18:07:00 (server time)

My transaction history last entries:

2013-08-08 18:08:07    DMS.SELLING    transfer-out (DeprivedPersonal)    30    0    0    0
2013-08-08 18:08:06    DMS.MINING    transfer-out (DeprivedPersonal)    30    0    0    0
2013-08-08 18:08:04    DMS.SELLING    transfer-out (DeprivedPersonal)    20    0    0    0
2013-08-08 18:08:03    DMS.MINING    transfer-out (DeprivedPersonal)    20    0    0    0
2013-08-08 18:07:00    DMS.PURCHASE    transfer-in (DeprivedPersonal)    30    0.02864    0    0.8592
2013-08-08 18:06:44    DMS.PURCHASE    transfer-in (DeprivedPersonal)    20    0.02864    0    0.5728

So we're basically there now - just got to test it a bit more thoroughly and make sure it handles things like network connection dropping etc.

We can't really know what MINING's true value is until after the DI has happened; right now all you can do is estimate. It's really going to depend on how many mining companies follow through with their Q3/Q4 deliveries. If Knc, ActM, Labcoin, etc come through, or even half come through, you're going to see some massive increases.

I'm tempted to buy MINING at .0025, which gives positive return with up to 40% increases each DI period, but there could be a massive jump come Sept/Oct/Nov this year...

It looks that MINING is now near it's true value.
Also we should really start marketing MINING towards those buying USB Block Erupters...

That sounds good! Much easier than stalking you on bitcointalk to see if you're online...