This is an extension of the article Basics of Risk Management. The stop-loss is the cornerstone of every risk management strategy. But it is also the trickiest part – learning how to place an ideal stop-loss requires a lot of effort.
The stop-loss was considered a meme in the crypto community. And understandably so – wrong stop-loss placement could shake you out from the fabled 10x run. Many people burned themselves and never touched stop-losses again. But when times are rough you can clearly see its benefits. If you were to continue using stops, developing and enhancing your strategy you would save yourself as much money as you lost missing an 10x.
My goal here is to provide a relatively simple guide to placing stops using different trading styles. I will cover most of the mistakes people tend to do using a stop-loss, manual and algorithmic stop-loss placement approaches and different types of trades and stop-losses techniques for each of them. Finally, I will cover situations when stop-loss is not an option.
Components of the Trade
Every trade is based on some thesis – price should move in X direction because of Y reasons. Risk management complicates things adding two more variables into an equation:
- price level P (take profit), which price should reach if Y reasons remain true
- price level S (stop-loss), which price should never reach if Y reasons remain true
An ideal stop will be triggered if and only if Y conditions are not met anymore. Then, all we need to do is to define a situation which will invalidate the reasoning of our initial thesis. Sounds simple in theory, but in practice it’s very troublesome.
A common situation is where a stop is placed too close to the entry price, and it results in the “shakeout”. Price reaches your stop-loss only to shake you out of your position, reverse and continue in the direction of your trade. It sucks, but it happens even to the best of us.
So in other words, our job is to define a stop-loss as a level which:
- Will be as close to the entry price as possible
- Once reached will indicate an invalidation of our entry thesis. It shouldn’t be reached just to stop us out, reverse, and continue trending in the direction of our trade
There are two faulty ways of placing stop: using hard stop and adjusting your stop-loss according to take-profit level to fit into your framework of wanted Risk / Reward ratio.
Hard stop or “fixed” stop is a stop-loss placed at a certain % below (or above) the entry price. For example, you could say, “I’m always entering a trade with 10% of my total portfolio, so if I place a stop-loss 20% below my entry-price, I remain in the bounds of 2% risk per trade, so it’s a good approach”. You couldn’t be more wrong.
Every setup is unique. Ideas behind different trades might be similar, but other things such as volatility, distance to resistance and support, market-maker activity and liquidity usually differ. More so, a stop-loss should be dependent on the time-frame of the trade – it doesn’t make sense to use the same stop of 5% on both 1H and 1D setups expecting great results.
Also, in theory, you can come up with some magic number which worked great in the past, for example, you backtest and find that stop-loss of 2.3% below or above the price works 90% of the time. But there is no guarantee that this rule will continue to work. Markets change all the time.
Sometimes you will find setups which have a Risk / Reward ratio lower than you usually trade. You might be completely sold on an idea that trade will work out – the only missing thing is that your R/R framework doesn’t allow you to enter the trade. In this option, a trader might move his stop-loss closer to the price, earning some extra R/R points.
This approach is wrong. Remember, we define a stop-loss as a level where our current trading idea is invalid. If you move it higher (or lower) just to fit into desired R/R you create a probability that you will get shaken out even if your initial trading thesis was right. You don’t want that.
If you find yourself in this position, just don’t enter the trade. Wait for the next opportunity. There will be plenty.
The opposite situation is when your target is so far away from your entry price that Risk / Reward is unrealistically high. It usually happens when you buy the support of some kind of consolidation pattern and expect price to break it up and continue developing new highs.
You might think that since your R/R is so high you can loosen your stop to provide some breathing room so trade can develop freely. But you would be wrong:
Imagine participating in this MTL/BTC trade. You spot a symmetrical triangle and price is simultaneously retesting lows of the triangle and previous consolidation level – you enter the trade expecting eventual breakout and retests of the historical highs.
You decide to put your initial stop-loss just below the previous swing low – you know that diagonal support lines are not that reliable and it can break below and still eventually go up.
Your R/R is 5.4 – it’s quite high and usually much more than required for the trade. According to the handy table from the previous article, this setup needs to have just a 15% winrate to keep you in profit. If you think that it’s an overkill you can move your target price a bit lower, for example to the next swing high.
But some people tend to do the opposite – they widen their stop. Here is one of the potential outcomes. Notice the stop-loss is much lower, at the next swing low – this way R/R is lower and you feel safe that price will have some room to reverse.
But what’s wrong with this way of thinking?
First, you enter the trade on the premise of the symmetrical triangle – the current stop-loss is way lower than the pattern. Even if the price was to reverse just above the stop-loss the triangle won’t be relevant anymore. We clearly broke down and the triangle starts to act as resistance – you can forget about your initial price target for a while.
Second, is there any reasoning why this area should act as the “reverse” zone? Is this level special? In this case, certainly not. It could continue falling just to reverse at the next swing low, or even at the historic lows. Additionally, wicks in this zone will trigger any stop you could potentially think of.
Finally you will waste time. Buying support is great because you enter at the optimal price – and once trade goes against you, you get an opportunity to exit fast and with minimal damage. In this case, you could even use a tighter stop-loss. The combination of the previous swing low & diagonal support is a strong zone, and the break of which indicates weakness. Or you could exit at the failure to break back in the triangle. It was a clear change of character and invalidation of the initial thesis.
Two Main Types of Trades
There are tons of trading styles, methodologies and technical indicators, but almost every trade breaks down to a simple idea: either the touch or the break of support/resistance level. These levels can be defined using swing highs/lows, price patterns, order blocks, moving averages or other indicators.
These two types of trades are completely different and require different position sizing and a stop-loss strategy. Remember, our task is to define a level at which the current trading idea becomes irrelevant – we look for the invalidation of the setup.
Support is a level which a price has a hard time breaking below. Usually, it’s an area with quite strong demand which pushes the price higher once it’s reached.
A good sign of strong support is an immediate reaction in the zone – there is so much demand that price has no chance of staying in it for a while. Resistance is the opposite – it’s the level an asset has a hard time breaking above. A lot of supply drives the price lower.
There are confirmed and unconfirmed S/R levels: a confirmed one is a level with at least two touches, and an unconfirmed one is a level with one or no touches at all.
Trading confirmed S/R might be appealing, but here is a twist: the more obvious an S/R level is, the higher a chance that there will be some kind of a shakeout of traders who try to play a bounce off this level. Additionally, each consecutive touch covers a part of the supply or the demand, so it eventually gets dried out and price breaks out. Also, once the level gets too obvious, there is a chance that price won’t even reach it – people who see it will place their orders a bit higher (or lower).
Let’s go over the example to see all the pitfalls of trading bounces.
Here is a BTC/USD chart with marked S/R levels. First, let’s talk about S/R.
$6700 – $6780 is a small area of consolidation inside a trading range which broke down and led to another rapid draw-down. Later, this level was retested with several long wicks and that confirmed its relevance as resistance. The support level is exactly the opposite – there was a consolidation which was broken up, and then repeatedly tested with sharp reactions.
Is there a trade? Not at the moment since we are in the middle of the trading range. But if we were to retest support or resistance, it would make a pretty good long or short setup. Let’s figure out our entries, targets, and stop-losses.
This last bounce of the support was meaningful: first, it broke the structure making a lower close, fooling many retailers that we go lower and then it immediately engulfed it with a massive buy candle. Usually, such structures act as S/R in the future, so we mark it as our potential entry zone. Our stop goes right below the wick of the previous swing low.
Possible mistake: you could place your stop below the wicks at $6200. It’s wrong because we enter on the premise of the most recent support level. $6200 could be a great support if the current one breaks down, and we will potentially look for longs there, but right now we are thinking only about the most recent information.
The same applies to the possible short play: we will enter somewhere in the resistance zone, placing our stop-loss just above it. We could move our stop even higher, to $6900 area (the highest point of resistance), but the principle is the same – we work with recent levels, right now there is no information about the relevance of the $6900 level. If the current setup fails, we might look for shorts there.
Eventually, a long trade plays out. We closed in the zone of resistance with a high entering the higher resistance zone. This zone is defined exactly as the lower one – a consolidation pattern, which later was broken down and now is touched only by a wick.
Possible mistake: a retail trader might be sold on the bullishness of the current setup – it looks like a text-book accumulation zone with a breakout upwards. But don’t forget two things: a) higher timeframe trend; b) overhead resistance. Entering longs on the breakouts in the downtrend is risky. And even riskier is entering a long right below another resistance. Keep it simple and play the bounces off the most recent zones.
Right now we are right under the resistance and it makes sense to look for shorts. To decide where to put an entry think of the highest level which most likely will be hit. The price might not reach highs of this long wick, it might remain within the bounds of the lower resistance. My best bet is to put it somewhere in the lower zone. The stop should go higher than the wick, and in this case, higher than the resistance zone to which this wick belongs to. If the price was to break above, the next level of resistance is vastly higher and we have no business sitting in our short.
Even though we didn’t enter at the best price, the trade worked out well and reached our target. The support zone turned up to be weak and it broke right through, but we couldn’t predict it without using order book analysis.
Since we broke right through the recent support and the trading range lows and we are in the bear trend, we might consider entering a short on this breakout. But breakouts will be covered in the next section.
Let’s take a look at higher timeframe action.
Generally, to increase the win probability of trades on the higher timeframe we should look for shorts (since we are in a bear market), so I skip longs for now. The first trade was a cleaner copy of the setup from the above – we draw a resistance zone using the engulfing candle which led to the break of the support. Later, this zone gets tested, forming a long rejection wick and we start to look for an entry in this zone. We place our stop right above the resistance zone (since a break would indicate a change of the structure and potential bullish reversal) and target at the closest support zone (don’t worry about catching a small part of the move downwards, you simply separate it in two different trades based on different setups: bounce of the resistance and retest of the broken support).
On the second trade, however, we got stopped out. We used absolutely the same thesis but this time we missed a move just by 1%. There is only one major difference with the previous setup – the lack of a rejection wick which indicates presence of the strong supply in this zone. It didn’t stop me from taking this trade in the real-time, and I don’t see a reason why I wouldn’t do it again.
But there is a lesson – higher timeframes require wider stops. You can nail the stop-loss and the take-profit on 1H chart with astonishing precision, but on higher timeframes – not so much. Most of the people look at 1D charts to get a feeling of the current structure and it leads to massive fakeouts and traps. Don’t get caught in it – if the only logical stop-loss level you see is very obvious, just skip the trade, or even better, wait for the fakeout and then enter the market.
The breakout is a breach of an important price level. Typically, a breakout of resistance is considered to be bullish and a breakout of a support is considered to be bearish. The stronger the level, the more power is required to break it. If a breakout occurs without the volume supporting it, it most likely will fail and the price will return to the range below the resistance.
There are three ways of trading breakouts. The first one is buying right at the moment of breakout, or in other words, buying without breakout confirmation. The second is waiting for a candle to close above resistance (or below support) before taking action. The first method is very sensitive – there is always a chance that current action is so-called “fakeout” and you will buy the exact top/sell the exact bottom.
Imagine using a breakout strategy and trading MEME/BTC pair. You wouldn’t be able to keep your sanity. It would be even worse if you didn’t wait for the closes for confirmation of the breakouts. On the other hand, buying support would be very profitable.
But in some cases, breakouts happen so fast that waiting for the close of the candle is quite costly.
If you were trading this BTC/USD daily chart and waiting for the close of the breakout to confirm it, you would pay 6.73% premium over the people who bought without confirmation.
So how does one define a stop-loss level whenever trading a breakout? Generally, breakout traders have a very high risk tolerance. Fakeouts occur all the time and they need to account for them in their Risk Management strategy.
It’s common to see a trader making several unsuccessful attempts before finally entering on the right breakout which covers all of the previous losses. This SUB/BTC example shows that, first, resistance levels are subjective and there are several ways to define them, and second, based on the chosen level you might get shaken out a couple of times.
Breakouts have a win rate way less than 50% – but they are still profitable to trade because successful breakouts provide very asymmetrical returns. It makes sense to use less risk % than usual. For example, if you are usually using 2% per trade, consider using 1.5% or 1%.
Since average win rate is so low, and we look for asymmetrical returns it’s not worth to enter breakouts of minor resistances – you want to have clear skies ahead.
SUB had several resistances above, with the strongest one around 5900 sats. If we estimate that average breakout win-rate is around 30% (from my own experience), then we need R/R = 2 or higher. Since our initial target is 7% higher than entry (or ~ 9% from the lower resistance), our stop shouldn’t be more than 3.5% (or 4.5%) below our entry price to stay at breakeven.
Let’s take a closer look at SUB/BTC breakouts. First of all, if a trader used the lower resistance line as his breakout level, the real win rate wouldn’t even come close to our 30% mark – he would make 7 losing trades before winning one, a total win rate of 12.5%. If he was to take every trade he wouldn’t be able to stay profitable even with a great risk management strategy.
But a good trader knows which breakouts are not worth trading. He looks at two things: volume and the potential stop-loss level. Sadly, volume on some of alts is non-factor since it’s usually inflated, but theoretical stop-losses would prevent us from entering trades #1 – #5.
Even before the breakout occurred, look for a logical level where you would place your stop in case of a breakout. Usually, this level will have either a swing high or low, some consolidation or a previous resistance or support which now should hold, or otherwise tells us about the change in the market structure. We wouldn’t enter trade #1 since potential stop-loss is way too wide: 6% below the entry price. Our R/R framework doesn’t allow to take this trade. Trades #2 – #5 are absolutely the same.
Trade #6 is questionable because a stop is exactly 3.5% below the price. On average, entering the trade won’t yield a lot of profit. Trade #7 is pretty good since stop is ~ 2.2% below the entry price. Unfortunately, we will get stopped out, but trade #8 will work out great. In total, we would lose 1R of our portfolio on the first trade, and make 3.82R on the second trade, resulting in a net profit of 2.82R. Not too bad, considering how bad it could be if we didn’t account for R/R deciding whether we should enter the trades.
Buying The Retest
The third method of trading breakouts is waiting for a retest of the broken resistance (or support). Once this level is reached, you enter a long (or short) position. This approach is by far the safest one: we won’t buy the exact top without the confirmation of the breakout and we get to look at volume – a great indicator that the breakout was significant. You don’t have this luxury when buying right at the breakout.
Another advantage of buying the retest is that you act as if you are buying at support – it’s common to see a level which previously acted as a resistance to become support after a breakout. Buying support is generally safer than entering on a breakout.
The main disadvantage is that not all breakouts continue with a retest.
Some of them don’t, and if you are a retest trader it’s better to let it go and wait for the next trade. Unfortunately, if such a breakout happens on a high timeframe, you might be up for a long wait. Also, the lack of a retest shows underlying strength behind the market. It often leads to parabolic price advance, so sometimes you’ll miss great trades.
Let’s come back to IOTA/BTC example from the above. The important thing to understand is that support and resistance are not lines but rather zones where price tends to bounce. But for R/R calculation you need to have exact prices of entry, stop-loss and target. A convenient approach is to use the middle of these zones for your calculations and entries.
We enter this trade on the retest of resistance (for convenience, in the middle of the area), and define target as the closest and strongest swing high. The stop-loss is as always the trickiest part. The best case scenario is when there is some kind of consolidation right below the resistance line – one might argue that once price breaks below it, the breakout invalidates, and most of the time it is the case. But IOTA went straight through the roof without stopping for a second. There isn’t a clear swing low or some other important level to consider as a stop-loss and we are forced to look for not so obvious zones.
There was an area (green) which acted as a resistance, then as support, then again as a resistance, and now potentially acts as a support in case price returns back into the consolidation area. In my opinion, if price breaks this support, then the case for a breakout is closed and we can safely move on without worrying that the price will reverse soon.
The problem is that this level is quite far away from our entry, and we lower our R/R ratio significantly. Potentially there is another level, higher than a current one which would invalidate breakout thesis. But unfortunately, price action doesn’t provide enough data to clearly see it. The only way is to use a “hard” stop several % below the entry, but as we know, it’s not reliable.
Price Discovery Breakouts
Move below an all-time high or below an all-time low is called price discovery – we are yet to see how price behaves at these levels. Most of the time people are scared of price discovery – it’s much easier to use price action when there are some obstacles and levels on the way of the price. We know what areas it should respect and what areas should be easy to breach.
There were tons of Bitcoin top callers during the whole run. People called top at every milestone – 3k, 5k, 7.5k, 10k. And you can’t blame them for being wrong – tops and bottoms are incredibly hard to predict. But still there are some factors which can clear the picture, you should consider hype, sentiment, media attention, market capitalization, fundamental analysis etc.
If some coin made a price discovery breakout, and you truly believe in its fundamentals and it could move north for quite some time, there is a great opportunity for a trade.
Here is a chart of ONT/BTC, a coin which did quite well. I would trade such a breakout if and only if I believed in the coin fundamentally and there was some hype – and not because I would be fine bagholding it.
Such trades provide the greatest Risk/Reward opportunities. You can use quite a wide stop-loss (for example, stops below the swing lows) if you believe that there is a lot of fuel to move price exponentially higher, so your R/R remains in the accepted bounds.
If you look at the price discovery of the another well-known coin, you can notice similarities.
EOS had a great run – there was a lot of hype and marketing, and many people considered this project to be a solid Ethereum killer. The similarity is in the lack of retests of the breakouts. Hyped up projects are usually very volatile and it’s easy to miss a breakout train.
It means that to have an edge you need to either understand lower timeframe price action quite well, or enter in anticipation of the breakout much earlier.
Even Bitcoin is treacherous when it comes to price discovery breakouts. The first one was unsuccessful, and the next one led to an explosive price movement killing all hopes of retest traders. The following one is tricky because the retest came much later. It was followed by a breakout with no retest, and the last one was the only breakout with reliable retest during the whole run.
One of the most well-known breakouts is ETH/USD triangle breakout. It was a perfect fit for our criteria: a fundamentally great project, enough hype, strong community and great potential for market-cap increase. We could enter right on the breakout or on the retest with several stop-loss levels for consideration. If, for example, you estimated that $1000 might be a great target for a number of reasons, then both of these levels would provide greats setups – higher stop-loss would yield a 15 R/R trade, and lower one a 5.5 R/R trade. Both are great depending on your risk profile!
One of the flaws of the manual approach is its complexity. There are a lot of variables to consider. An outcome of your strategy is based on your subjective interpretation of the chart and your discipline and skill to remain within bounds of your trading plan.
Some of the traders prefer to leave themselves out of the equation – they build objective trading systems, rules of which are set in stone. If this system underperforms, they tweak it, backtest it and put it back to work. Instead of manually validating every trading setup they continuously look for issues in their system and make small changes.
Since most of the time rules of such trading systems are binary – enter the trade if X condition is met, put stop-loss at Y level, exit if stop-loss is reached or Z conditions are met – some traders go one step further and create or purchase trading bots. This way they don’t need to spend a lot of time watching charts and setting alerts. Instead, they can focus on the enhancement of their trading strategy or fundamental knowledge of the market.
In some sense, this approach is similar to investing, but instead of investing in a specific market you invest into your trading algorithm. It will require some time to provide a clear picture whether your algorithm is paying off and if it requires some changes.
If such method attracts you, keep reading! I will cover some of the most popular indicators providing a trailing stop-loss. You may choose one that suits you or create your own.
Average True Range
Average True Range is an indicator which measures volatility of the market. The range of day’s trading is simply high – low.
Welles Wilder (creator of RSI) introduced the concept of True Range as the greatest of the following:
- Current high minus current low
- Current high minus previous close (on an absolute basis)
- Current low minus previous close (on an absolute basis)
Average True Range is an N-day smoothed moving average of the True Range values. Wilder recommended using 14 as a parameter of the moving average.
ATR = [(Prior ATR x (N – 1)) + Current TR] / N (with N = 14 by default)
So how does ATR serve us as a stop-loss indicator? We’ve talked about the shortcomings of hard stop-losses. You may tweak your hard stop, moving it 0.1% higher or lower until you are happy with your backtest results, BUT any method based on some static parameter is bound to eventually fail. Markets are never static – they change all the time. If after countless sleepless nights of backtesting you found a golden rule of “stop-loss 1.268% below or above the entry price” which was working like a charm for the previous year, there is still no guarantee that it will work for the next year or even the next month.
If a hard stop is not an option, then how does one decide what stop to use? It makes sense to use wider stop-losses in volatile markets and narrower stops in calm markets. This way you won’t get shaken out of your position when the market is moving and won’t lose more than necessary when the market is flat. ATR helps us to do exactly that.
ATR % Stop Method
ATR is common indicator for defining initial stop-loss. Usually it is placed at a multiple of ATR below the price (if long) or above the price (if short).
A common multiplier is 2, but depending on the market it may vary. Backtesting might show that another value is preferable.
Let’s take a look at this BTC short setup. Imagine that our fictional trading system told us to short at the close below the local support ($6640). Where would our stop go? First, look at the last ATR value – in this case it is 37.2. Then, multiply it by a parameter of your choice (I used 2) and add it to the price of the entry: 6640 + 2 * 37.2 = 6714.4. Round it up however you find suitable and you are ready to go!
Note that our initial stop was very close to the entry price – there was next to no volatility in the previous hours. After we entered the trade price rapidly decreased and ATR reacted the way it should – volatility increase caused ATR to grow.
If we were to follow our next long breakout signal we would need to use a much wider stop: ATR is 61, so our stop will be $61 * 2 = $121 above or below the price depending on the direction of the trade.
ATR Trailing Stop
As we already know, ATR is a great tool to determine initial stop-losses accounting for the current market volatility. But it is also useful to build a trailing stop-loss. ATR can be used to create bands around the price. Basically, each candle has a band above and below it – current price plus or minus current ATR multiplied by some value.
There are several custom indicators on TradingView which do just that. The one I used is called CM ATR Stops/Bands. It allows to set not only a moving average period and multiplier but also a timeframe – for example, you can refer to 1D ATR using any other timeframe.
On this 4H chart I’ve used 1D ATR as values for the band. As a result ATR band values change periodically, every six candles. Even if the day was quite volatile there are some 4H candles with no movement at all – I don’t want them to affect the bands and cause false exits. Also, I’ve used 20 as a moving average period and 2 as a multiplier.
The idea of a trailing stop-loss is simple – whenever you are in long and a bottom band moves higher – you move your stop. The same applies to the short and upper bands moving lower. Eventually, you will get out of position once the price touches your stop-loss level.
The main rule of trailing stop-loss is to never move it lower (if long) or higher (if short).
To demonstrate this rule, look at the chart above. In the second trade, in which we went short, we moved our stop-loss four times – each time when upper band value decreased. Out last stop was around $6700. As you can see, next day’s ATR changed and the value of upper band increased even before we got stopped out. If we were to move our stop higher, we wouldn’t get stopped out, and in fact, we will rarely get stopped out at all. If the market was to enter a bull phase, the price would go up and so would the bands causing us to move stops higher and higher until we are out of profit.
Remember, trailing stop-loss is there to help you secure most profit possible in case of the market reversal.
Chandelier Trailing Stop
Chandelier Trailing Stops are also dynamic since they are based on Wilder’s ATR. Chandelier stop is N-period ATR added (or subtracted) to the N-day high (or low).
Chandelier (long) = N-day High – ATR(N) x Multiplier
Chandelier (short) = N-day Low + ATR(N) x Multiplier
An indicator from the chart above is called “Chandelier Stop” and is available for free. It uses a 22-period and 3-multiplier as default settings and was able to catch Ethereum trends pretty well during backtesting.
The main advantage over ATR trailing stop-loss is that it’s more stable. It will drastically change only if there was a significant shift in the volatility or creation of higher highs and lower lows. It means that it might be more suitable for traders who wish to trade longer-term swings. I’ve found it useful on 1D and 4H timeframes.
The idea of the trailing stop-loss and the main rule remains the same – you don’t move stops in the opposite direction of the entry even if the indicator tells you to do so. When short, move stops only lower. When long, move stops only higher. No exceptions.
Parabolic Stop And Reverse is another creation of J. Welles Wilder. It’s a very popular indicator for trend identification but is also used as a handy way of determining a trailing stop-loss.
SAR follows the price. If it’s below the price, the trend is supposedly bullish. Above the price – the market is bearish. Name of the indicator implies that it works best in the trending markets, which usually end in some kind of parabolic move.
It’s especially useful in altcoin markets where it’s common to see such a parabolic advance. On both charts from the above Parabolic SAR provided great exits. But everything comes at a cost – during a ranging market this indicator is pretty mediocre at best and you might want to use something else. But in my experience, ranging markets are death for almost any trend-following indicator.
So what is so unique? How is it built? Calculations are too complex to put into an overview article, but the main change is in addition of special parameter – Acceleration Factor. This factor has a starting value of 0.02 and is increased by 0.02 each time there is a new highest high or lowest low in the current trend. Each new SAR point is built by multiplying the previous one by this Acceleration Factor and some other variables.
In the end, if the trend is strong, SAR points move higher (or lower) at a very fast pace, implying that for the trend to continue price should increase parabolically. If it doesn’t, one might argue that the trend is about to change.
SAR is brilliant because it accounts not only for the change in the price over a predetermined period but also for the time – J. Welles Wilder knew that once price rapidly increases, it basically puts a timer on the market – if price fails to continue its acceleration, change in the trend is inevitable.
Wilder concluded that trending markets occur roughly 30% of the time. Obviously, it changes from sector to sector and cryptocurrencies might be different. But nevertheless, most of the time markets are choppy. In my opinion, SAR is inefficient for trend identification but is amongst the best automatic stop-loss indicators in the trending markets.
Bill Williams’ Fractals
Bill Williams’ Fractal is a series of five consecutive candles forming a specific structure: the highest high should be surrounded by two lower highs for buy fractal and the lowest low should be surrounded by two higher lows for sell fractal.
In some sense, Bill Williams’ Fractals are swing highs and lows. They provide different trading opportunities, including breakout trading, but are also used as trailing stop-losses.
Bill Williams built a ton of different indicators which work very well together and form a holistic trading system – I’ve covered it in my other article. Be sure to check it out!
Entering a long (short) initial stop goes at a level of the last sell (buy) fractal. Then, with the trend progression once new fractal is printed it becomes a current stop-loss. The main rule remains the same – move stops only in the direction of the taken trade: lower for shorts and higher for longs.
In this BTC/USD example, we took a trade based on the break of the short-term support, placed an initial stop at the last buy fractal and moved it 3 times on each subsequent lower fractal. Luckily, we got stopped out right before the massive bullish candle.
It’s also an option to use the second last fractal as initial and move stops accordingly. This way you may create a bigger room for price fluctuations. It’s up to preference and requires backtesting, but usually, on smaller timeframes (1H and below) it’s a suggested method.
The Bill Williams method is one of my favorites because it’s based on pure price action – fractal is an important structure, the break of which usually implies that something has changed. Same way swing highs and swing lows do -it works like a charm on smaller time frames.
Managing Risk of Speculative Investments
Speculative investment is a short to medium-term investment which may be based not solely on the undervaluation of an asset but on the speculative basis – market cycles, market maker activity, hype, flavor of the month coins, ICOs, etc. I don’t include long-term investment in Bitcoin and other blue-chip cryptocurrencies in this category.
What separates such investments from basic trades is the expected return – usually it starts with a couple of hundred % and goes up to infinity. The drawback of such investments is that usually using a stop-loss is not the best idea.
Why? First of all, sometimes it’s not even an option. For example, when investing in ICOs, you might have to wait some time before coins will get distributed (look at Tezos for reference), and even once they get distributed there might not be a market for them to trade. Also, there are always other risks such as scams, security flaws and changes in regulation. There is no guarantee that you will ever get to sell your ICO coins even for a loss.
Another common trade approach is buying an asset not because it’s fundamentally solid but because you spot some kind of market making activity which indicates that it has potential to pump.
Huge wicks, random volume spikes, low satoshi cost, everything shows that something is going on.
And eventually, from the low of 85 sats price skyrocketed to 10412 sats – an astonishing 12200% return. But everything comes at a cost – spotting an opportunity there would be quite a challenge, and an even harder challenge is taking it.
The main concern is liquidity – there were days with 0 volume. No trades at all. Other days had an average volume of 0.01 BTC. First, how do you get in? The only way is to place orders all the way down and waiting for them to be hit. It will take some time, but worse, you expose yourself to the risk of holding funds on the shady exchange.
Second, how do you get out? Well, if it goes the wrong way you simply don’t.
You bet that if this coin pumps it will create enough liquidity for you to exit. But when it goes down, there is no buying pressure at all. RKC from the chart above went from 2000 sats to 100 sats in just two days.
So how to approach the risk management in this case? Let’s go over conditions again: if you are right, your efforts will be well rewarded; but if you are wrong – you probably won’t be able to exit without a substantial loss.
It makes sense to invest only what you are willing to lose.
To apply a Risk Management framework you need to accept that your stop-loss is basically at 0. Then, since you take the risk that in case of failure you lose all your funds your maximum position size is equal to some fixed % of your portfolio.
The great example of this approach is the @needacoin strategy. He takes it to the next level – whenever he wants to enter a coin, he looks for a position size of 1-2% of his total balance (so he manages up to 100 positions). Then, he enters the market gradually, on the way down, in the 20-25% partials of the chosen position size. Once a bag is filled, it’s a waiting game.
In some way, this strategy is similar to angel investing. You accept that most of your coins will die a horrible death, but ones that do not will cover all the losses. It takes just one 10x to cover 10 failed investments.
If you made it this far – thank you for your attention! I hope you’ve learned something new about stop-losses and are ready to apply your knowledge in trading!
Since the article is quite large, let’s go over the main ideas again:
Every trade is based on some thesis: price should go in X direction because of Y reasons. A stop-loss should placed on a price level, which once reached indicates that these Y reasons are not relevant anymore. Additionally, stop-loss should be as close to the entry price as possible.
Stop-losses should be dynamic and unique to every trade; you shouldn’t use a “hard” stop or place it somewhere just to fit into your R/R framework.
Any trade is based on a break or touch of some level: a support or a resistance. Bounces are more reliable – if there is no external pressure (such as massive market buying or selling) price should remain above support / below resistance. But if the breakout is solid, then possible returns of the trade are much higher.
- Bounces – higher probability, lower profitability if it works out
- Breaks – lower probability, higher profitability if it works out
It makes sense to use less % fixed risk when trading breakouts and higher % when trading bounces.
Support and resistance is not just a price level but rather a zone. Stop-loss should go above or below this zone or you are risking getting stopped out frequently.
Be careful trading higher timeframes. You might get used to the precision of the trades on low timeframes (sometimes you can nail the trade to the dollar), but on higher timeframes shakeouts and fakeouts are much more extreme. Use wider stops.
For people who don’t enjoy manual trading and prefer bots and algorithms there is plenty of stop-loss indicators out there. Check out ATR, Chandellier, Parabolic SAR and Wyckoff’ Fractals.
Sometimes using a stop is not an option. For example, ICOs lock up your funds for quite some time and the final market price is out of your control. Low-cap speculative altcoins are so dry on liquidity that you probably won’t be able to exit with a stop-loss. The only way to exit if liquidity catches up – and it happens only when a coin shows a decent pump. It’s an all or nothing situation. Invest only what you are willing to lose.
Thanks again for reading. Good luck in the markets!