Topic: Goomboo's Journal - page 25. (Read 281459 times)

To bring us back full-circle to the question which has relentlessly haunted this journal: should I systematically trade or buy and hold?  Examine the chart below and then let's discuss.



By all metrics, trading the crossover has outperformed buy and hold since the day the crash began.  Trading has returned 134% and buy and hold has lost 37%.  Traders experienced a 68% drawdown in value, investors watched 82% erode away.

Even though the metrics seem to say that trading has "worked" better than buy and hold, it's still pointless to make the comparison.  After all, when you're asking a question of "what works", you by definition must select a timeframe to make a judgement.  The last two months have been favorable for traders but the last 6 months have rewarded investors.  No matter which side of the aisle you stand, investor or trader, you can find supporting figures for your viewpoint.  There are only two ways to "settle" the debate: price instantly rises to $1,000,000 per coin and stays there forever or price slowly and indefinitely falls to $0.00.  In the meantime, it will remain impossible to definitively state which mentality is "better".

This remains an ideological question.  Are you "voting with your fiat" and supporting an idea, or are you here to earn a profit?  It is my viewpoint that disciplined trading is the most logical method of approaching the markets.  The purpose of this journal is to convey what it takes to profitably trade.

In order to provide full disclosure for the original system, I have backtested the 10/21 hourly crossover from the day of the market crash until now.  The system has continued to grind out consistent profits...with important caveats.

• There have been 34 trades closed since April 9th of which 35% were winners and 65% were losers
• The average winner has been 2.9 times the average loser
• Cumulative return of 134% before commission and assuming no slippage

If you are interested in being a trader, you must critically analyze the above performance.  Specifically, you should take a hard look at the first two weeks of April.  During April, the system experienced a reversal of a 31% profit to a 37% loss.  This means that if you were following the system, you would have watched 68% of your account value erase over the span of two weeks.  This brings three very important points to the table.

• It is essential that you have a plan to get out of the market when price roars against you
• A method of trading less size during periods of drawdown is important for surviving in this business
• To fully capture the profit potential of trend following, you should seriously consider only trading markets which easily allow buying and shorting

Absolutely.  Losses are the name of the game unfortunately.

Edit - in case you're wondering, I do agree with you - it isn't the "best" example.  There are better, cleaner charts to be found.  However, I stand by what I said in my post - those who followed through with the system should have profited from that downturn.  It is entirely possible that these profits were clouded by commissions / whipsaw action, but that unfortunately is the nature of the game.

Best of luck with your trading.

Nowhere, which is why if one is to accept the 7% "protection" (now officially 4.7% - without the fees), he must also have endured the previous losses.
Of course, you never said it was a 7% of pure profit, but singling out this trade was, IMO, a little disingenuous.

An average of the last 50 or 100 trades would be more representative.


Which is why I'm saying it's not a good example.
It was a trade that could look good in your trading record, but it had nothing to do with trend following, which is the only goal of your EMAs crossing technique.

Again, I'm not arguing against the technique, but rather the representation you gave by using the last drop.

Where in this thread do you see me advocating a percentage threshold for participation?



Who cares?  Do you think this point of contention will matter one bit in 24 hours?  A week?  In the long run, the method tells the same story.


Re-read my post.  I think you're reading a little too much into what was said.

Yup, but this is assuming an immediate sell on crossover (without waiting for any additional % for confirmation).
This means that while you might have saved ~7% on this drop, you would have done at least 9 other transactions in the graph you are showing, completely falling behind that 7%.
(Which you did not mention when presenting the glorious 7%.)

And what if the drop would have occurred 4 hours before it did? The first candle, which is still the bulk of the drop, could have easily occurred while the crossover was telling "buy".


In summary, my beef with this example:

• You single out the 7% out of context to show a win, while the last days (weeks?) are clearly losing.
  (Which is kinda odd, since you took the time, at multiple occasion, to show the importance of long term results.)
• You took an example that had nothing to do with trend following, which is the sole point of your EMAs selling/buying technique.
  Any further drop could be associated with hourly trend following and that would make an excellent example.

Again, I'm not disputing the technique, but the example chosen.

Even if you sold at the bottom of the second hourly candle after the dip, you could still have a decent profit.  Best case, after the first candle was solidified, you could have sold at 130 in the next hour.

For systematic trading, there is no "kinda".  You are either following the signal, or your are relying on your gut/intuition.


You handle it by making a point to aggressively manage risk.  As shown through backtest, trend following methods win 30-40% of the time.  This means that the majority of the time, you will be thrashed by market noise.  It's your job as a trader to make sure that the trashing you will receive damages your account balance as little as possible.  How risk management looks varies from trader to trader but a general guideline is to decrease your trading size during periods of loss.

Some people also try and establish criteria to filter out the noise such as "only trade the crossover after price makes a new N-day high or low".  In my experience, these methods typically do not hold up as well as the original system in backtest.

PS, I see you're relying on trendlines for support.  Just a little advice a successful day-trader once conveyed: trendlines are better drawn with a Crayon than a pencil and they tend to be broken on the 4th touch.  So basically, by the time most people have drawn trendlines and prepare their trade, the market blows up in their face.

It's pretty much a universal truth of trading - if everyone's doing it, everyone's going to lose money.

Welcome to BTC trading - this is precisely why I do not call BTC a "professional" trading market.  I'd wager there are fewer than 5 individuals actually living from BTC trading profits.  The spread (while it has decreased over the past few months), commissions, and slippage for large traders is just killer.


I'm not here to display cherry-picked charts which woo participants.  It's a great example because it's the nasty truth of trading - there are rarely any good examples.  And ironically, most of the good examples are the trades that the majority of participants weren't in until the end because they threw in the towel after a string of losses.

In fact, I'd be willing to bet that the majority of people reading this thread will not capture even 60% of the profits provided by the original system in the next "great example" due simply to a lack of emotional courage.

PS I'm not sure exactly what you're looking at, but here's where you should have sold using the hourly system from the opening post of this thread:

Yes.

Are you using hourly candles in that chart?

Yeah, kinda.

While an immediate sell on crossover would have protected you from a 7% drop, it would have cost you a fortune in fees and slippage in the last few days:



I do appreciate the knowledge Goomboo is transmitting, but I think this is not a good example.

Also, the EMAs did not recross. The price could go back up before the next crossing.

Goomboo is correct. Kinda..



Great thread by the way. I think this is a good example and sorry, if it was asked before.

MA cross trading is great, if you have a clear trend, but depending on the timeframe, you had multiple crossings in the last days. How do you handle sideway movements?

Picture from Bitstamp.

I have nothing to say about the trading bots - I provided no input for any of them.  I'm speaking strictly about trading the trend, as measured by the system taught in this thread.  Followers of the system should have had their assets protected from a 7% drop in value over the last 18 hours.

Not on my chart - the signal was generated 4 hours prior to the sell-off.

If anything, the recent selloff suggests the use of a trading bot isn't such a good idea after all. I was lucky enough to turn off my bot the night before and wake up to the large drop. Manually doing things is probably the best way to go from here.

I don't like this example.

Yes it should have been a surprise (if following the EMAs): the selloff is mostly what caused the crossing.

For those of you trading the more active timeframes (Hourly / 4-Hour bars), this recent sell-off should not have been a surprise.

The disciplined trader maintains focus during times of volatility to profit from the herd.

Yeah, I'm running Tobli's bot, v0.2.2.3.  It seems to work well, although that 4k dump last night was quick enough that the bot only sold near the bottom before the market bounced back up.

xx78213, I was using a similar bot (Tobli's fork) and it had the same problem. Frozen. I just went to the source and updated it again and it all works now. Perhaps you could try updating your bot?