I appreciate your point of view, and completely agree that you are chasing the trade and will most likely not be filled (big issue with stop-losses). I'm actually agreeing with you right now, there isn't much use for a stop loss during a big and fast crash like that.
Thanks for sharing your thoughts, nice chat!
To continue it... lets make some money.
Earlier this morning we talked a bit about what was working & what was not, but I am interested in using you guys as a sounding board to get deeper into the mechanics of _why_ Bitcoin is a different market then anything I have worked with before, and honing a strategy to maximize profits for our little group here.
As stated before, I do not believe it "trade magic" or secrets, it has all been tried. Anybody who ever tells you they have a secret system is delusional or trying to sell you something. Apologies to the very fine folk at Butter, but you don't need their bot to use EMA and profit from it; it is one of the easier systems to use manually.
That said, Butter is the most valuable single tool I now have in my arsenal. Not because of the trade bot (which I have already stated is the best one I have found), but because of the back-testing engine.
In the past, I would spend weeks with a grease-board charts designing and testing systems, trade points and theories. I would come up with a better trading strategy in my mind and have to spend dozens of hours proving or disproving it's worth using historical data. Butter's ability to do this in seconds is like finding the Thor's Hammer of trading.
So, back to the hot topic of the day. If there are still those out there that are still trying to jump out of the market the instant they see trouble coming, it is because you have not tried it enough times in the real world. Especially in the Bitcoin market, by the time you see an actual flash-crash (sure, let's call it that), it is too late to sell. You are chasing the landslide and you _will_ lose a ton of money trying to catch it.
Here is an exercise to sharpen your skills. With a good charting system, go back one year with GOX trades. Identify and mark all the flash-crashes. Now, go back and count off all the crashes that telegraphed a true trend reversal verses those that did not. 67% of the time, a downward trend followed the FC. That means 43% of the time the market corrected and continued upward.
If you sold during the slide, you lost your shirt 43% of the time, because the market corrected, continued up, and you had to buy back in higher then you sold; often a lot higher. This because you sold while chasing the market down, and then had to buy while chasing the market back up.
Now, back at your chart, let's look at the 67% of the time the crash did actually telegraph a trend reversal and selling was appropriate. Take the lowest value 7 minutes after the first long red candle. In my experience, this is where a quick acting trader or bot can "catch up" to a fast falling market. Remember, just because set a bot to dump @ 5% does not mean that is where it will trade. You (or your bot) have to find buyers in a panicking market. This is always going to be well below where you would like it to be. Mark this number.
Now, move forward on the chart 45 minutes to an hour. See the bounce back? There are very few things that "always" happen in trading, but this is one. There is always a correction after unusual pressure. I tell students in to think of these extreme downward movements as having a rubber band attached. Because this pressure is finite, there is always a snap-back before the trend continues. _This_ is where you want to sell, _if_ you have identified a downward trend. Now, instead of chasing a market that is running away from you, you are selling into a market that is actually pushing up into your sale. Timed right, you will get a much better price and a quick sale.
The larger the amounts you are trading, the more important this technique becomes. Good luck selling >$20,000 worth of points to a fleeing market at a price you can live with. You want that market moving hard into your sale to efficiently unload such a position.
If you are accomplishing these trades manually, timing and patience is the key. It is the hardest thing in the world to watch your stake fall through the floor, and resist selling. That is why everyone else is chasing the trade down. But you did your homework. You know that _every_ true panic-sell results in a hard, if brief, correction. _Wait for it..._ lol.
If you are trying to program a bot to catch the snap-back, we get to hone down the requirements to just timing (because computer can be programmed to be patient). We can do some very simple math and find the average timing of all the snap-backs in the last year. Remarkably, I am finding this timing to be very consistent. I have found that selling 1 - 2 hours _after_ the crash is identified virtually always results in getting a better price in the sale then if I had tried chasing the slide.
Finally, there is the element of bots. There are a _lot_ of bots in this market. It is estimated that because of the techie nature of this market, the bots actually number in the tens of thousands. Scalping bots, arbitrage bots, RSI bots and limit bots are all reacting to increasingly slighter twitches of the market. Bots are trying to fool other bots into reacting in order to profit off of the swings it causes. Yes, I am playing this game with them (mostly with STB right now), but although it is fun, there is much less actual profit in it when compared to good, old-fashioned solid trading.
So, full circle as to what we really want to see added to this bot.
Stop-loss? For me, no. it defies the trading strategy that by nature is the best stop-loss already; if set right.
Profit checking? Some have asked that the bot make sure it is not making a losing trade. Again,this defies the principals of this type of trading. This system will make trades for loses specifically because it then always has you in the most favorable market position. If it failed to get me into the market because the trade would have resulted in a small loss, and the market then gains 100 points, I would be very unhappy indeed. In fact, the most recent huge gain was immediately predicated by a trade resulting in a small loss... to get me back into the market. The "profit checking" suggested would have negated this gain. No place for that in this type of trading.
Magic wand to tell us exactly when we are at the top or bottom of a trend? Yes please. Lol. But since there is no magic in trading, I am very satisfied with a good, solid machine that based on good numbers and research returned a damn-near 50% profit in the last 30 days.
