Whoops, I forgot to update yesterday. It was another hold day.
Today, however, I'm liquidating my long position, at a loss.
My position just before liquidation:
BTCUSD 18.0 $5.6715 -$1.60 -1.563%
My account status currently:
USD $102.38714
Topic: The flipist method - page 5. (Read 13969 times)
You win (kinda). According to my calculations here, a flipist method that closes its position out at the daily close versus a trading strategy that uses the EMA10/21 crossover and closes its position at the daily close loses only $4.0892 compared to $-45.00281.
Interesting. I seem to recall Goomboo posting the results of backtesting they'd done for the 10/21 EMA crossover method, with rather phenomenal profits. Have to go take a look to see why the difference is so huge to your test.
If I am doing it correctly then it is probably that people are a little more selective. It looked like they used hourly time scales but I may be wrong. On daily time scales the EMA10 and 21 only crossed once for the six month period I looked at. For hourly data there was enough crosses to go short 82 times and long 82 times. My model also assumes that the position was liquidated at the end of the trading day. Most currency traders are not going to wait to take their profit when a sell signal is given. By that time they have lost their unrealized profits. They will have a stops to take their profits.
You win (kinda). According to my calculations here, a flipist method that closes its position out at the daily close versus a trading strategy that uses the EMA10/21 crossover and closes its position at the daily close loses only $4.0892 compared to $-45.00281.
Interesting. I seem to recall Goomboo posting the results of backtesting they'd done for the 10/21 EMA crossover method, with rather phenomenal profits. Have to go take a look to see why the difference is so huge to your test.
Today I do nothing!
Here's where we're at currently:
BTCUSD 18.0 $5.6715 -$2.27 -2.224%
Here's where we're at currently:
BTCUSD 18.0 $5.6715 -$2.27 -2.224%
Aand I'm in again!
BTCUSD 18.0 $5.6715 -$1.79 -1.752%
Long! Going up guys!
Or maybe the powers of randomness are displeased with me. We shall see! I'll now go do something random to gain favour.
BTCUSD 18.0 $5.6715 -$1.79 -1.752%
Long! Going up guys!
Or maybe the powers of randomness are displeased with me. We shall see! I'll now go do something random to gain favour.
You win (kinda). According to my calculations here, a flipist method that closes its position out at the daily close versus a trading strategy that uses the EMA10/21 crossover and closes its position at the daily close loses only $4.0892 compared to $-45.00281.
Aand I'm in again!
BTCUSD 18.0 $5.6715 -$1.79 -1.752%
Long! Going up guys!
Or maybe the powers of randomness are displeased with me. We shall see! I'll now go do something random to gain favour.
BTCUSD 18.0 $5.6715 -$1.79 -1.752%
Long! Going up guys!
Or maybe the powers of randomness are displeased with me. We shall see! I'll now go do something random to gain favour.
And the third 2 in a row! Maybe I should start taking bets on the results of the flip...
And again, I am doing nothing today, as much as my fingers itch to press a button... Not that I know which one I'd press, but fortunately I don't have to worry about that kind of thing!
I guess this means the price will go down. Or up. Or maybe it'll stay about the same for the next 24 hours.
I guess this means the price will go down. Or up. Or maybe it'll stay about the same for the next 24 hours.
Technically, it's not a coin. To be clear, I'm using random.org to generate a number between 1 and 2 to decide what to do.
Today, it's another 2, so I'm still holding. Yesterday was neutral, it seems - the numbers said to hold at no position, but didn't say to short, so I didn't take any losses, but neither did I profit. Perhaps the subatomic particles aren't pleased enough with me yet, but they don't seem terribly displeased, either.
Today, it's another 2, so I'm still holding. Yesterday was neutral, it seems - the numbers said to hold at no position, but didn't say to short, so I didn't take any losses, but neither did I profit. Perhaps the subatomic particles aren't pleased enough with me yet, but they don't seem terribly displeased, either.
Why am I not subscribed to this thread yet? I love it 
Again, there's no fucking way this is a coin...
Hey if you want to sell me that coin, that'd be great
Hey if you want to sell me that coin, that'd be great
LOLO
i only read the first post, but: this is madness!
I LIKE IT!
i only read the first post, but: this is madness!
I LIKE IT!
The flip says to hold off from doing anything today, I guess that means the sideways movement will continue!
OK. Back to trading now that I'm at a computer again. The flips today had me liquidate my long position at a profit of 0.456%, but told me not to go short. Back to uncertaintyville then!
My balance now:103.98USD
My balance now:103.98USD
i'm calling bs... this coin is right more frequently than anyone else
I ran two simulations over the mtgoxUSD price from Jan 1, 2011 to Dec, 31 2011. One simulation gave me a rate of return of about 100%, another gave me a rate of about -300%. When I get some more time I will bootstrap this model to get a better estimate.
this coin is obviously the manipulator just toying with us all - for a heads up
i'm calling bs... this coin is right more frequently than anyone else
Nothing doing today. It's a bull market... At least according to random choice.
I may not be able to perform these trades this weekend, will have to see.
Oh yes. Almost forgot to update. Seems we're in bull mode now, so I bought 15 BTC on Bitcoinica @ 6.2(ish).
I'm considering moving my trading to Gox or something, though, I had a hell of a time getting the buy to even go through today thanks to the lack of reserve. I'll reimburse Zhou's 25 bucks if I do that, of course.
I'm considering moving my trading to Gox or something, though, I had a hell of a time getting the buy to even go through today thanks to the lack of reserve. I'll reimburse Zhou's 25 bucks if I do that, of course.
You need to account for time of the day. Randomize that as well. Also, I would use multiple coins. If you want historical data pm me.
I'll look into randomizing the time for trading if I get around to figuring out how to use Bitcoinica's API and automate this trading.
What do you mean by multiple coins?
The coin you use could be biased. Maybe part of the coin is worn down leaving one side heavier than the other. To truly test for stochasticity there should also be 1000 people tossing the coin. I will run a simulation within the week and post my results here.
Basically what I will do is take 5 dice using random.org's algorithm. If 3 out of 5 of the dice are even numbers then that is buy bitcoins with USD, if 3 out of the 5 dice are odd then sell bitcoins for USD. If a buy position is already set then further buys will be a hold position. If a sell position is already set then further sells will be a hold position.
Keep up the experiment. I love stochastic processes. It is what I work on in my day job.
Jump to: