Topic: I want to trust one of you who tells me, buy now and sell now. Nothing more. (Read 6824 times)

Thanks for sharing your strategy.

As for Gox dump, I'm not sure. Those who wanted fiat badly already've paid the 10% premium and left. The rest want to leave only because they can't. It's kind of Catch-22. If you can leave Gox, why should you?

So, if problems resolved soon, no dump because Gox isn't a problem anymore. If not soon, no dump because Gox won't matter anymore

Thanks for your validation Rampion. I'm also preparing fiat, and also plan to stop the fixed allocation to the upside, meaning I plan to not sell anymore if it goes up, and only buy more if it goes down. In this case if the bottom happened already and it is mostly up from here I am protected.

In my opinion the *logic* thing would be BTC going to the '60s again before resuming the long-term uptrend, but as I said many times this market is hard to read accurately because its extremely small. Add to that the problems Gox is suffering regarding withdrawals, and there you have a big question mark. In my opinion, *if* and *when* Gox solves its problems for good, there will be dumping - especially if those problems are solved suddenly. If they pull their shit together gradually (reducing the withdrawal queue, whatever), I won't expect big crashes because of that. Will see.

I can share with you my current strategy: I'm holding my coins but I have some fiat ready to buy more if there is a big dip. My main bet is that Bitcoin will consistently go up in the long term, following the historical trend, so I increased my investment funds and decreases a lot my trading funds - in fact I'm using the little I have left for trading just to try some arbitraging (and failing big time because of Gox's delays). I'm not so sure about what will happen in the short term, so I play cautious not risking coins but preparing fiat to not miss a good chance to buy cheap.

I wish I knew  I'm not a guru, I just happen to know a bit of calculus.

With stock market it's reasonable to assume return to pre-bubble line, because stock's fundamentals are not affected by a media hype. But bitcoin is different: it's fundamentals are in network effect, so the hype should improve it. So my guess is that the line of after-bubble exponential grow should be above of the pre-bubble line. We may be already on the line and then the return to 60's won't happen. But again, it's just a guess and your is just as good as mine.

Obviously b&h worked really well for you, so I want to make it clear that I'm not trying to convince you to use a different method. But there are some points in your argument where you're assuming too much about the alternative methods.

re: "Robot like trading". Goomboo says he does, but I don't do it. In my view, there's a scale, from "pure intuition" to "purely mechanical" trading. Following, roughly, a strategy like the ones in Goomboo's thread simply means to lean strongly towards the mechanical side, but that doesn't mean you never use your intuition.

re: Goomboo's original method. The thread has substantially evolved. Goomboo's original method (1 hour EMA20+10, no threshold) is completely *unprofitable* now. That's why the thread has been about finding and backtesting different parameters, for the past 30 pages or so, and they have been found and posted.

re: Time commitment. Sure, an hourly EMA method takes more time than b&h. But you should keep that argument separate from the profitability question. If you say (near) algorithmic trading is too time consuming, that's a personal choice. If you say it's not profitable, that's something that can be (somewhat) tested.

re: Your own trading method. I remember you sold your trading stash at around 130 in May, correct? (good call, by the way :D). What made you do that -- Intutition? TA? Probably a combination of the two. My point is, not everyone has an equally strong intuition, especially not new investors. There are several parameter combinations of a *daily* EMA crossover that would have given a sell signal in May as well (a bit later than 130 though), and would have resulted in a nice increase of your coin stash when buying back in July. Note that in this case, it wouldn't have been very time consuming (because of *daily* EMA), and not very expensive wrt fees & slippage (because only very few trades).

In the end, everyone uses the method he likes best, too many personal factors play a role here to find an "objectively" optimal method. Many other forum members tend to recommend buy & hold as the "default" strategy, but I wouldn't.

In my opinion, the best new, relatively inexperienced investors can do is to employ a "modified buy & hold" strategy. Which means to, 1) use some TA to decide when to buy in, 2) use some more TA to decide when it is really just better to sell (like in May), and in the meantime, 3) just hold.

Algorithmic strategies *with a comparably low sensitivity* (like daily EMA crossover) can play a crucial role in deciding on 1) and 2) -- unrelated to the questions your raise about how high fees & slippage are, or if you should follow TA indicators mechanically or not.

Great post Rampion. I think this is the best critical analyses I've read about Gomboo trend following analyses.

I'm afraid I have to agree. Thanks for sharing Rampion.

Thanks

Do you think deflating bubble period/sideways period is over or do you think this will take another half year?

My calculations say that it should be over only by oktober/november (since exp growth took 3 months, and correction has always been 2.5 times longer) and a low of $60 should be approached again then since the long term exponent growth will be at that level then.

But I also feel the fear I could be wrong and by then it might be $300 never to come back.

Would value your opinion.

You are welcome. It's simple math. For real-life results you can, as Rampion recommended, test it on historical data.

As a consolation, it's devastating only in periods of exponential growth. First months after the bubble, when it deflates or goes sidewise, it's OK. But later or, when fundamentals catch up and the exponent resumes ...

OMG that makes me see how devastating this fixed allocation is when it goes up

Are you sure it is square root? I never did simulation on this I must admit. I was already wondering though howcome my total capital grew so little after having been right on a tenbagger.

Thank you so much Wary for showing me the math.

Yeah, nevermind any of that stuff I wrote above. Just took a 4% hit due to bitstamp slippage. Not my best day :/

When I'm daytrading I also consider EMA crossings, but my point is that doing it systematically (which is the whole point of using a system ) is much inferior to b&h, and in fact is unprofitable many (if not most of) times. Some of Goombo's posts suggests the opposite, but then he admits he is not following his own system

I think this is your course of action: you check the EMAs and use that info to decide whether to trade or not, but you factor other elements (market depth, sentiment, overbought/oversold indicators, money flow, etc. etc. etc.). Plus, you place limit orders (as I do 99% of the times), so you probably miss quite a lot trades because the order is not met after the cross signal was triggered. During a down/up move, a wall as small as 500BTC in the opposite direction of the trend will be probably eaten just a little, if at all. Breaking it in smaller walls spreaded on a bigger price range is also not very effective when the price is going in the opposite direction . At the end of the day, that's very time consuming and pretty much like gambling, there are a lot of factors to consider, and it is in no way the cold, systematic, robot-like approach Goombo seems to suggest in his thread.

Following my experience, the time consumed in analyzing all the different indicators to daytrade (should I wait for a bid wall to raise so I can sell into it? are we overbought? how close are we from the 50/100/whatever day EMA? etc. etc. etc.) is usually not worth it. This is not Forex, the market is so small that a single whale buy/dump invalidates all the technical analysis you did by taking the price many points up or down, regardless of what technicals were saying. I've personally increased my BTC "trading funds" by aprox. 50% in the last months, but honestly I invested too much time on daytrading, and I was able to have that result just because it was too obvious that the price was going down after a bubble pop. Before the April, 10th crash I just held, and I think it was a good choice. I'm now giving a try to arbitrage (waiting +10 days for a SEPA transfer coming from Gox), but my "daytrading" time is finished for now. I increased my coins during an obvious opportunity (post-bubble crash), now in my opinion is time to hold again (at least for me, taken into consideration my average entry point is in the $10-20 range). I'm not saying price won't go sub $100 again (I think $100 will be tested again sooner or later), just that for me its not time any more to risk coins by daytrading unless some other obvious situation happens.

Well, I use a crossover strategy like that (not really, but sort of, let's say) and I do place limit orders. In fact, the order placing is the one part where intuition really comes into play with this: waiting for a wall to appear (but obviously not too long), waiting for a mini-retracement of the price to buy/sell into, things like that. I don't have quite the volume you mention, but not that far from it, and I actually trade on bitstamp.

Note also that, at least on bitstamp, slippage when selling can be even higher than your example, but when buying, it's near zero on that exchange On average then, it's not that bad.

Hourly EMA crossings are not an option for me either, at least not with the usual parameters. What I do is look at recent history, decide after the fact which swings I would have liked to trade, and adjust my parameters accordingly. It usually runs down to 2 to 3 trades per month, depending on volatility.

I guess I see your point: mechanically following an EMA strategy that is also fast is not an option. I agree. But there are a) slower methods based on averages that can be traded pretty much mechanically, and b) you can take the average based method signal as your main input, and then make the final decision based on additional indicators/methods, and then execute the trade with a bit of craftsmanship. The latter is what I do, and I'm doing much better now than when I traded purely intuitively when I arrived here.

Absolutely right about fees and slippage being inversely proportional. In fact I'd consider for a backtest between 0.2% and 0.3% fee (which is still CRAZY high by the way) but a much higher slippage (+1%). Seriously, market depth at Gox (lets leave alone Bitstamp) is usually a joke, I have had 3% slippage (!!!) in the past by selling something like 500 BTC which is a modest amount for a professional trader. That was quite an extreme situation, but nevertheless +1% slippage is completely normal with medium volume. And that's a lot when trading hourly EMA crossings.

Your only choice to avoid slippage is to either a) place a limit order, which is obviously totally ineffective when trading EMA crossings; or b) waiting for a bid wall big enough to sell into it, which is again out of the question if your strategy is to systematically trade EMA crossings.

Just a very quick reply: a) I edited my post to make more clear that I do think b&h is a good strategy for most investors, especially those who are not interested in putting in a lot of effort, but even for them, using some TA to determine the entry point for b&h is IMO extremely useful.

b) fees and slippage was the starting point of my little test series:  Most people in the goomboo thread set trading cost to 0.4%. I started from the assumption that  those costs were  too low,. I decided on a fixed cost of 1%, estimating 0.5% fees and 0.5% slippage, per trade(*) -- I'm sure you will agree that's pretty costly, and greatly favors b&h which has zero costs.

(*) fees and slippage are roughly inverseley proportional, I'm sure you agree: higher volume means lowers fees, on all exchanges I know, but more slippage, and vice versa.

Thanks for your interesting (as usual) insight, just a question: did you factor a) slippage and b) trading fees in your backtests? What kind of trading volume did you consider in your backtests?

Following my experience Goombo's strategy is inferior to b&h when taking into consideration slippage and fees, especially if you have a medium/high volume per trade as the slippage is HUGE as soon as you sell/buy a few hundred coins. In fact, Goombo's backtests are for hourly 10/21 EMA crossings, but then he admits that he does not trade on that timeframe - he trades weekly or daily at most, precisely because he has to hand-pick very seldom which "cross signals" to trade, as otherwise slippage and fees will eat all his profit. In my opinion that kills his system's purpose.

At the end of the day, he writes a lot about emotionless trading based on EMA crossings, but he admits he is not using a bot (the only way to be 100% emotionless) because it would be unprofitable (again: fees and slippage + a lot of false signals), so he has to "hand pick" daily or weekly when to trade. In fact, he never specified what "strategy" he follows to choose which crossing to trade and which not, which leads me to think that his choices are not based on a precise system as it looks like and are somewhat arbitrarily chosen taking into consideration diverse factors...

At the end of the day he is making an "informed gamble" as all the others TA daytraders do.

As the most simple strategy, b&h is probably not bad advice for newcomers, but I would still advise against unconditional b&h, for two reasons:

(1) a realistic estimation will see the number of new investors correlate with price and volume, i.e. more new people entered the market in April than now, for example. Agreed so far? The newcomers who bought in close to the April peak are still sitting on a loss -- even though they will turn a profit the long run in all likelihood. I suspect however that quite a few of them left the market again in frustration.

My conclusion: I can only see benefits from teaching newcomers a modicum of TA to determine if conditions are severely overbought and about to correct, and if price is extremely volatile and it would be better to wait before buying in. Using very "conservative" tools, a new market participant whose money arrived on the exchange in April should have probably waited until early May, then bought in at a reasonable 100 to 120 USD, compared to buying in immediately at 160 to 200, or even more in the worst cases.

tl;dr assuming bitcoin continues its growth path, it ultimately doesn't matter at which price you buy in, but for very understandable human reasons, it actually matters a lot. Some TA can help with finding the right entry point -- afterwards, b&h all you want :D

(2) Buy and hold is outperformed by several parameter combinations of a simple average crossover strategy -- I backtested this myself. There is reasonable doubt about this of course: the parameters were optimized on the trading period itself, which is problematic. I have a long post saved that I'll post in goomboo's thread at some point, where I backtest several parameter combinations that outperform b&h under a more realistic methodology.

What I'm trying to say is: I'm pretty sure that careful use of relatively simple (and testable) TA methods will allow some people to outperform b&h reliably. Not everyone perhaps, as it requires a bit of patience and willingness to get technical -- if you simply try to apply a recipe, you're probably going to lose.

EDIT: added a sentence or two

@RationalSpeculator: I have the impression you might not like the "buy&hold" strategy because you want to make a lot of money with Bitcoin, and you believe that by investing "only what you can afford to lose" you won't radically change your life, because the amount you can afford to lose is too small. Thus, you throw to BTC money you can't comfortably risk to lose, and that's why you embrace an historically losing strategy as fixed allocation.

Honest question: I am a getting close? I would appreciate an honest reply.

My opinion on the above, is that your "fixed allocation" strategy is driven by greed at the end of the day - you want to make a lot of money ASAP, thus you throw at BTC a too high % of your total wealth. You want to make a lot, so you invest more that you can lose, thus you need to mitigate the risks with your strategy.

Let me tell you that IMO the above would be an emotion-driven decision, and thus a huge mistake.

You speak about $1,000 like being a remarkable milestone - let me tell you that it's very short sighted. You seem like a thoughtful guy, so I'm sure you have analyzed the forums. Didn't you read what people said about $10 in 2010? It seemed completely crazy. And $100 in 2012? Only some permabulls (like Adam) wrote that $100 would have happened in 2013, and they were laughed at.

$1,000 its around the corner. If the Winklevils ETF is accepted, you will see 5 figures per BTC before you can even blink your eye. You know the very characteristics of BTC, its unique properties, the real problems it solves... And that there will only ever be 21 millions, right?

Just lurk the forums: in 2011, a lots of guys with thousands of BTC cashed out for good during the bubble inflation/deflation. They posted about how they bought a nice house, how they made a killing... But now in 2013 we know they did a mistake. They made a few hundred of k's denominated in fiat - now they would be millionaires, probably "all set for life". But yeah, hindsight is 20/20 - in 2011 nobody really knew if BTC was going to fully recover from the bubble pop, the sentiment after the crash was nefarious, the beating on the media continuos and the "this is a ponzi" crowd might had a point.

In 2013, there is no excuse. You know how the market performed in the past, you know what's the historical BTC trend, you know that the ones who held on their coins bought or mined sub-dollar are the ones winning big time.

Do you think BTC is just another commodity? Or on the opposite you think there is much more in BTC, that is a revolutionary tool that could profoundly change the world, an experiment that will probably just skyrocket or go bust? If you believe the latter, there's no better advice than "invest all the money you can comfortably lose". And if you are a trader, trade to increase your BTC stash and forget about short-sighted fiat profits, because you will probably regret it in the future - as all those who "secured profits" in 2011 are doing now.

Yes. I do this with some of the BTC. But I like  tinkering with the rest...

This. Say, if you keep it 50:50, then your capital will grow twice as slow. That is when BTC will grow 4-fold, your holding will grow 2-fold only. When 100-fold, your capital will grow 10-fold. Square root. In Rational Speculator's case (20% in BTC) it would be fift root: 2.5-fold instead of 100