
The Art of Trading Currencies, the World's Largest Market (w/ Dave Floyd and Brent Donnelly)
What this covers
This Real Vision Live conversation pairs David Floyd, founder of Aspen Trading Group, with Brent Donnelly, a senior HSBC FX trader and author of The Art of Currency Trading. Floyd mostly asks the questions, and Donnelly draws on his career, from Lehman Brothers in 2005 to 2020, to describe how a professional approaches currencies. Much of the discussion is about process: how he sizes trades, places stops, and keeps records. Donnelly says his results come out at roughly half winning days and half losing days, so a year's profitability depends on the ratio of average up-day gains to down-day losses. He also describes how the market maker's job has changed as algorithms took over routine tasks.
The middle of the conversation covers the tools. Donnelly treats technical analysis as a way to define where you are wrong rather than to forecast. He is dismissive of Fibonacci and Elliott Wave, and prefers simple moving averages and round-number levels, which he says matter because real institutional orders sit there. He also describes "slingshot" reversals, where a watched level is briefly broken and then reclaimed. Other topics include why traders should automate exits instead of trusting themselves once in a position, why they should take profit at the target, and a rare kind of panic illiquidity when one side of the market vanishes.
The macro and market-reading material is where the two diverge slightly. Donnelly warns against reflexive contrarianism and favors a selective version tied to a narrative that has run out of steam. He illustrates this with the 2015 euro decline, when bearish targets kept being extended. He argues that ECB negative rates backfire and that attention has shifted to German fiscal stimulus. Floyd prefers trading with the trend and adds a more contested point about the intuitive "feel" that comes from watching one market for years. Donnelly closes on specializing in a few markets and on the unglamorous work of reading central bank speeches.
Donnelly argues that successful currency trading requires combining macro understanding with disciplined micro execution through position sizing, technical analysis for risk management rather than forecasting, and deep specialization in a few markets—not simple systems or trend-following alone.
- Technical analysis is necessary but insufficient for forecasting; its real value is in risk management and tactics
- Position sizing and win/loss ratio discipline determine long-term profitability more than any single trading signal
- Deep expertise in 3-5 specialized markets beats broad generalist knowledge because it enables reading order flow and narrative shifts
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Every trade should have a predetermined stop-loss to avoid catastrophic account-blowing losses, and traders should automate stops and take-profits at night (not actively monitoring) because once a position is on, physiological and emotional changes occur in the brain that impair judgment, making it dangerous to trust yourself to make rational exit decisions in the moment.
“in my rules of trading rule number one don't blow up uh avoid risk of ruin above all else so yeah personally i every single trade that i put on i have a stop loss i have a scratch pad in front of me as soon as i put a trade on i write down what my stop is and i automate it so i have a system where you know if these trades do this and every single trade that i put on i always have a stop loss and at night i tried to not obsessively check the market so i always leave a stop and a take profit every single night so whatever positions i have um i always have a stop and i always have a take profit”
Successful discretionary trading fundamentally requires rigid, systematic position sizing discipline because traders naturally tend toward one of two extremes—either sizing too small due to hesitation and fear, or sizing too large due to overconfidence—and without a mechanical rule (e.g., fixed percentage of capital per trade), emotional biases prevent proper execution even when the underlying trading methodology is sound.
“i mean the hardest part of discretionary trading is you can develop whatever process and however robust that process is there's generally with two extremes either you tend to size too small because you're hesitant or you tend to size too big because this feels like a huge opportunity and everything feels like a five-star trade i tend to be more on that side of the continuum where like i tend to get a little bit too excited”
The goal of a trading career is not to be right on every trade or find the 'holy grail' system that guarantees profits, but rather to develop a sound methodology, apply it with discipline, manage risk obsessively, and accept that consistent long-term profitability comes from getting fundamentals right, sizing correctly, and adapting as conditions change—not from a simple system.
“so i think the reality is that a lot more has to go into your process than just technicals but without technicals i don't think you can really succeed any either so again going to the idea of like necessary but not sufficient condition for success”
Simple technical levels like support and resistance work not because of mystical chart properties but because large institutional orders are placed at round numbers and psychological levels by patient, long-term oriented portfolio managers who are willing to wait for good entry points, creating actual order book density that other traders observe and trade around, generating self-reinforcing price action.
“a lot of times you can explain why something very simple like support and resistance works is because there's huge orders in the market and those people that are putting those orders in the market are patient and they have a medium-term or long-term time horizon so say somebody says i want to be short the end and i want to buy dollars but i'm not really that worried about it and i i want to get a good entry point so i'm going to put a bid for 2 billion at 108.50 and you know if we get down there maybe the rest of my portfolio is doing well so i don't need to panic to buy it and then suddenly every time it goes to 10850 it seems to bounce off of that level um and then people start noticing it and then all the you know bids start forming above and stops below that level”
A market maker's role has evolved due to automation: while algorithmic systems now handle routine tasks, the human role has shifted to executing large trades and serving as a subject-matter expert who helps clients strategize around their trades, while positioning one's own book to anticipate short-term market moves without being caught wrong-footed when client flow arrives.
“the market making role is pretty broad and it's been kind of dynamic over the years because of automation so a lot of market making now in all the businesses at banks is automated so what that leaves us to do is execute the big tickets and then also the role much more has become to be the expert in in your currency so like whatever currencies you trade you need to be the expert in those things and then you can communicate with clients and help them strategize”
Specialization in 3-5 products within one universe (e.g., G10 FX) produces an edge that trading broader or more diverse assets cannot match, because deep expertise enables traders to read all relevant information (central bank speeches, economic data, policy reactions, order flow patterns), whereas generalist approaches across multiple asset classes or markets dilute expertise and rarely generate profitable insight.
“at various times have gone more like touristy as people would call it and then i've always come back to for me it's g10 fx is the thing that i'm good at and um whenever i've deviated i've learned a lot of stuff but it has never been tremendously profitable so like me trading oil or gold or or things like that or s ps being probably the most difficult thing in the world to trade i've always come back to g10 fx and there's a guy that i know that used to work at a hedge fund and when he started they said okay the only thing you're allowed to trade is norway so you can trade norwegian rates no region equities and that's it and his um career path really benefited from that because he understood he flew to norway and like met the guys from the central bank and stuff”
Reflexive contrarianism—automatically betting against whatever is popular—is a losing strategy because trends are durable (positioning follows price), and the human tendency to trend-follow is strong, so being contrarian just to feel smart is fighting the market; instead, selective contrarianism based on analytical reasoning about when a narrative has expired is more effective.
“one mistake that i made a lot um when i was young because like it feels great to feel smart because if you're the if everyone's short and you're long and you make money you feel super smart but it's a really bad way to trade um so like reflexive contrarianism is not a good idea because you're just basically fighting every single trend because generally human beings follow like you know positioning follows price so if something's going down people want to sell it and things trend a lot so um being reflexively contrarian is the worst the worst i would never advise people to be contrarian all the time”
Very occasionally (roughly once every few years), a currency market will experience extreme illiquidity where one side (bid or offer) completely disappears, signaling panic conditions, and these panic moments are typically the best time to take profits or exit positions because panic always eventually subsides and represents an execution edge.
“it's actually interesting what you said about the market making thing because very very rarely um like maybe once every three years um there'll be a time when the ecn in a currency doesn't have a bid or an offer like one or the other like one time euro norway was absolutely going crazy and no one knew why and there was a bid and there was no offer and weirdly that is the few times like i've maybe seen it about 10 times in 20 years it's always the absolute extreme so it's like the the absolute panic and i think looking for those panic markers is actually a good strategy like generally panic subsides you know and and so those intraday periods of madness if you're getting to your take profit you just take it and be happy and if it keeps going who cares you know you made your money”
Technical analysis should primarily be used for risk management and tactics (i.e., defining when you are wrong and where to exit) rather than as a forecasting tool for predicting market direction, because very few successful traders use charts alone, charting patterns are easy to automate and therefore get arbitraged away, and empirical research shows that back-tested chart patterns do not remain profitable out-of-sample.
“generally technical analysis should be used for tactics and it should be used for risk management and a lot less for forecasting where the market's going to go and there's a bunch of reasons for that number one reason is very few to actually zero traders that i've ever worked with that were very successful only use charts um another reason is that it's that stuff is the easiest stuff to automate”
Sentiment analysis as a trading tool requires direct access to market participants—if you work at a bank you talk to institutional clients, if you're independent you monitor social media and networks, but you're looking for signs of extreme positioning or anger/conviction (e.g., angry emails in response to a contrarian call), not just headlines or aggregate sentiment indices which lag actual shifts.
“if i put out a bullish s p call i'm saying if you do um do i get a lot of angry emails about it um if yes and that is an indicator you know so the indicators aren't just looking at the cftc data um which i'm personally not a huge fan of um just because it's very slow moving um but i think there's a lot of ways to get sentiment whether you know if you work at a bank obviously you talk to a lot of clients if in your case you talk to your clients um but if you don't have clients you have twitter and um yeah you have uh linkedin and all those kind of places and you have friends to talk to and you can get a sense of okay the sentiment is extreme now”
To trade successfully, one must understand the current macro equilibrium and then anticipate which new information or headline will shift the market to a different equilibrium, meaning traders who understand the fundamental context can quickly assess which news is meaningful (market-moving) versus noise, avoiding the trap of reacting mechanically to every headline.
“the way that i generally look at the world is you have to understand the macro but if you're trading short term what you're trying to do is identify okay that currently we're in this equilibrium what's going to push us to a different equilibrium and and how much are we going to move and can i anticipate that before it happens and so some of that is is simply if you understand everything about the current macro environment and a headline comes out then you can react in a in an efficient way to that and know like okay this is meaningless this is not meaningless”
Slingshot reversals—where a major price level that many traders are watching (e.g., a round number like 100 yen) gets taken out briefly, then reverses back below it—represent high-probability setup because the breakout of the level triggers stops and forced buying/selling, then the reversal exhausts that energy and leaves positions exposed, making the reversal setup more reliable than following the initial break.
“the best setup i feel like is when you take out the level and then you go back below it um so i feel like at that point there's so much energy that builds up around the significant reference points or like major major levels that a lot of people are watching that i feel like you take all the energy out and then anyone who is short has stopped out anyone that wanted to do anything has basically done it and then i feel like it's a lot safer to be short”
When a position is profitable and you've reached your initial take-profit target, you should take the profit rather than hoping for a larger move, because the psychological bias to hold becomes overwhelming once you're in the money and you feel overconfident—humans systematically underestimate how lucky they were to hit that initial target and overestimate how much more they can make.
“it's really easy to trade badly on good trades because you're kind of in the money and you're feeling good and you feel overconfident so i try to automate as much as possible um both on the stop-loss side and the take profit side”
Traders should collect detailed P&L data—either by individual trade or by regular time-based snapshots—because such data reveals process leaks and inefficiencies that traders may consciously or subconsciously know but are reluctant to admit, allowing targeted improvements, and simpler performance metrics like win rate and winning-day-to-losing-day ratio often reveal more actionable insights than complex measures like Sortino or Sharpe ratios.
“if you collect a lot of data on your on your trading um i think you can learn a lot and i again like i always kind of err on the side of simplicity like i i'm familiar with sortino ratios and and like sharp is it pretty simple but um like the more complex ways of analyzing p l and for me i always come back to the simpler stuff of like your wind days versus your down days is that ratio good is it improving is it worsening um you know what percentage of the days do you win i know some traders can win on 60 of the days um so like that is a very personal thing it depends on how you trade and so but i think knowledge of that and collecting that data can and really help you improve your process and just see little leaks and you go okay well that's that's a leak and now how can i plug that and sometimes the leaks are pretty easy to plug once you know about them”
Simple moving average-based indicators (e.g., 20-day and 100-day or 5-day and 20-day) are more useful than complex esoteric methods because they fit trends well, their breaks signal momentum loss, and they may have algorithmic support (large TWAP trades maintain 45-degree angles until they complete), making them both simple and mechanistically grounded in actual market structure.
“so if you just throw a moving average onto a trend in in fx like say you say you take an hourly um euro dollar chart and you just put a couple of simple moving averages on there like a 20 and 100 a lot of times it'll fit the trend really well um and so what i like to do is just essentially it's kind of like drawing a trend line but fit the tr fit the moving average to the current trend and then it just gives you like a simple thing of like okay whatever was driving this thing lower has lost a little bit of its power for for whatever reason”
Simple trading ideas are easy to market and sell, but the reality of trading is that developing an edge requires years of dedicated learning, with most traders ultimately failing even after extensive study.
“i think the idea of something simple working is very appealing and also very unrealistic so it's easy to sell um like online uh a package of technical you know holy grail kind of things but it's a lot harder to sell like hey if you do a whole bunch of work and spend 10 years learning about this business then maybe you might be able to make money even though most people don't make money you know what i mean”
Understanding central bank reaction functions—the specific policies and concerns that drive policymaker behavior—requires reading numerous speeches and policy documents, and this work is necessary but often unglamorous and boring, yet absolutely required for any trader specializing in currency markets.
“there's a lot of speeches and a lot of like boring stuff that you have to read to be keep abreast of what the central banks are thinking um but to me that's like absolutely necessary like if you're not willing to read some boring speeches from the bank of canada you can't really understand their reaction function”
A trader's long-term success is determined almost entirely by the ratio of winning days to losing days, not by the accuracy of individual trades or the sophistication of the trading approach, and Donnelly's historical data from Lehman Brothers (2005) through 2020 shows a consistent pattern: he wins approximately 50% of days and loses approximately 50%, with yearly profitability determined solely by the ratio of average up-day profit to average down-day loss.
“i know from many years of like i create a new spreadsheet every year and it has my p l that i pretty much win 50 of days and i lose 50 percent of days and even like i used to work at lehman brothers back in 2005 and i have those spreadsheets and it's it's kind of insane how the consistent that is yes what is determining my success in any given year is the ratio of my up days to my down days that's almost the entire determinant of of a period for me”
Watching a market for extended periods develops an intuitive 'feel' for price action and likely movements that is difficult to articulate but represents genuine expertise, similar to pattern recognition that firefighters or attorneys develop through years of experience.
“you know there's been studies done on more of the um the neurological aspect of it but you know the longer you watch something whether it be you know you're becoming a you're becoming an expert attorney or a firefighter or whatever you get this inane sense or this experience that you can call upon and sometimes it just happens where you're like you know i know it's time to go along the euro here but you don't logically know why that's the case but your brain does have a way of interpreting all this vast amount of data”
ECB negative interest rates are counterproductive for monetary stimulus because they harm the banking system and don't stimulate economic growth as intended, so market participants have shifted focus toward waiting for German fiscal stimulus as the next potential game-changing policy lever.
“people have kind of come to the conclusion that the ecb can't do too much more because negative rates are hurting the banking system and maybe they don't really stimulate anything anyways so the panacea now is german fiscal”
In 2015, the euro collapsed from 135 to 120 to 110 because the ECB was conducting QE and corporate issuers were selling euros to fund operations, but the market narrative kept extending bearish targets (135→120→110→95) as the decline continued, until eventually the corporate selling dried up and ECB sentiment shifted, causing the narrative to lose credibility.
“this was true in 2015 ecb was doing qe and all the corporates were issuing in euros and euro was collapsing and you know when euro was at 105 everyone thought when it was going from 135 to 120 people's targets was 120 then they moved their targets to 110 then it was at 105 so everyone moved their targets to 95. and then the narrative started changing as like the the corporate selling kind of dr dried up at that time and then the ecb wasn't quite as dovish and it kind of seemed like maybe they were getting a little people were getting a little bit annoyed at how much the currency had moved”
Esoteric technical analysis methods like Fibonacci retracements and Elliott Wave are unreliable because after a market has moved substantially, one can draw so many overlapping Fibonacci levels that the chart resembles a Los Angeles highway interchange full of lines, making it easy to retroactively justify any price action but providing no real predictive power.
“if you take um a market that has moved a lot and then start drawing all the fibos on there like you can basically fill the thing so it looks like a los angeles highway interchange where there's just lines everywhere and part of the reason that i say that is that you experience it a lot where people go hey that this is the 38th one of this move and then you go look and you're like oh i'm looking at a different move and then all of a sudden there's like six different 38 ones”
Trading in the direction of the overall trend is more consistently profitable than contrarian trading, and Floyd prefers letting the market prove an upside move before attempting to catch a bottom in a downtrend.
“i think so many people are they're so tempted to want to call the bottom or the top because one if you do it boy does it feel good yeah but obviously your your chance of doing that is usually like 20 to 10 or 15 percent it's very low but i do think there's a time and a place to place those trades but it's usually few and far between i think trading in the overall direction of the trend tends to be at least for me tends to be the the better approach”
Moving average parameters (e.g., 100 vs 200 vs 252) matter less than adapting the parameter to match the market's actual trend speed, and over-optimizing parameters through backtesting produces results unlikely to hold out-of-sample.
“i feel like mostly it's not that important what parameters you pick you're like like i know when i remember reading john where facebook you know way way back in the day and you know there's a lot of talk about how you can optimize things and stuff i feel like you know whether you're using a 100 day or 200 or 252 because that's how many trading days there are in a year i feel like it's more like you adapt the parameters to the market and you know if you see that like in euro sweden the 200 hour always seems to match the trend then okay i guess that's kind of the speed that that euro sweden moves at”
The Bank of Canada's reaction function shifted over time: in 2016, policymakers were extremely concerned about the oil price collapse and its impact, causing oil-CAD correlation to be very high; after oil panic subsided, they became more focused on real estate risk in their rate-hiking decisions.
“the bank of canada is actually a good example because their reaction function has changed a lot over the years so they were extremely worried about the oil collapse in 2016 for example and governor poloz spoke a lot about that in all his speeches and in the meetings so you knew that the correlation between oil and and canada was going to be really high and then after a while once that sort of energy mini energy panic dissipated they became more concerned about real estate so then you know as they're hiking rates into a heavily indebted economy you know that they're going to always have one eye on the housing market”
Technical analysis functions similarly to knowing opening poker hands: it is a necessary condition for success but insufficient on its own, and like common technical patterns, poker hand knowledge can be understood by novices but does not generate long-term profit if it's the only edge.
“it's kind of like if if i'm sure a lot of your viewers play poker it's kind of like knowing the opening hands in poker um like it's a necessary condition but it's not sufficient and that's what i feel like with technical analysis is that it gives you some indicators and it tells you okay if this breaks i'm wrong but as a forecasting tool i just don't think that it's it's enough”
Coronavirus headlines gradually declined in market impact because initial visceral fear dissipated; traders who understood this evolution could better predict which coronavirus-related news would move markets and which would not.
“a good example of that is how the coronavirus headlines have slowly decayed in terms of their importance because the visceral fear in the market has dissipated and so after a while you understand that that it takes a little bit more to move the needle”
Identifying shifts in market narratives—when the consensus explanation for price movement is losing credibility and a new narrative is emerging—is more valuable and profitable than most other trading approaches, particularly when the market is heavily positioned on an old narrative that has lost legitimacy.
“the main thing that i think i'm good at is understanding the current narrative and then identifying like oh something's changing here like there's a new narrative trying to sneak in and obviously there's all these cl narratives clashing all the time so there's a skill in understanding okay this thing actually matters and this thing doesn't um and then if you combine so say say there's been a long ongoing narrative of um i mean this isn't true but let's just say the market's been selling euros for six months and everyone thinks”
The narrative shift from virus fear to stimulus expectations in 2020 was an identifiable future shift that traders could potentially profit from, by being long currencies like AUD after waiting for technical confirmation of the transition.
“that's a good example i think of a future narrative shift that probably will happen and that people are trying to time right now is the negative impact of the virus versus the positive impact of the stimulus which will follow so it's actually a decent example i think because it's it's a kind of narrative shift that you can maybe catch and then you know be long aussie for for the big china stimulus but that's a huge if and it's not happening right now”
To anticipate short-term FX market movements, traders use multiple overlapping tools: technical analysis, fundamental analysis, sentiment analysis, cross-market lead-lag relationships, and these same techniques are used by retail traders and by institutional market makers.
“in order to do that you use a lot of the same tools that anyone would use who's trying to generate alpha which is you know whatever systems you use uh and you know in terms of technicals fundamentals sentiment cross-market uh lead lag all that kind of stuff which i'm sure we can dig into a little bit later but so you end up using all of those techniques that somebody sitting at home trading would use similar techniques i would hope to to anticipate short-term movements in the market”
Book recommendation: 'The Hour Between Dog and Wolf' explains the physiological changes that occur once a trader takes a position, showing why emotional decisions in winning positions are inevitable without automation.
“there's a great book about this um called the hour between dog and wolf yeah i'm sure you've read that one yeah the second that you put a trade on there's physiological changes that are happening in your body”
Most trading books are marginal in quality, but Donnelly's 'Art of Currency Trading' stands out because it reflects genuine practitioner methodology rather than theory, evidenced by his methodical process and practical focus.
“I've been in this business for 20 plus years most of the trading books out there are let's be nice in terms of my word choices marginal um when i read this one i knew that brent was actually a practitioner of what he did because of the way it was written in the way he spoke you could tell that his process was very methodical”
Donnelly positions his book as a guide to developing personal trading style and philosophy: readers should take insights that resonate and ignore the rest, rather than viewing it as a prescriptive system.
“you know take what what resonates with you and and ignore the rest i think everyone has to develop their own style and you know i read a lot of books when i was coming up trading and a lot of it i didn't like and some of it i liked and you take what you like and you throw out the rest”