You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Global Macro Trading Journal
- Thread starter Daal
- Start date
Quote from Daal:
The trading books talk about fixed % in order to size your positions. Like 1% or so. Seems totally wrong to me. They are not adjusting by trade quality and EVEN IF THEY ARE, the ratios of adjustment(say 2% for avg trade and 4% for a good trade) is not based on anything mathematically rigorous, why the 100% jump?Why not 150%?
With Optimal F you can get numbers that are based on things that are actually sound and mathematically proven
Fixed position size makes more sense for mechanical trading approaches, where conviction levels and trade profiles don't vary from trade to trade. For a trader who can do everything from swing trade to spread to buy and hold for deep value, there is no one size fits all.
I've never found a really great book on position sizing. Some of the market wizards interviews were helpful in a "here's a concept to kickstart your research" type of way. Van Tharp gave some food for thought. In terms of risk management tools, Ken Grant's "Trading Risk" is excellent.
In my opinion, this is one of those areas where you really have to know the deep-level concepts to get a good handle on things. Either that, or accept a suboptimal compromise of overly rigid rules.
Just my .02
Quote from Martinghoul:
Not really valid to compare GBPDEM and EURCHF floors, for obvious reasons. Moreover, who's to say that 9bn is all they got? They could be on the bid in 100bn or 500bn tomorrow, assuming they still have full political support from all quarters.
I agree it's comparing apples and oranges, but thought it was important to get the reasoning of those on the other side of this trade out there, and there is another side. There are some who think 100% of traders are long EURCHF.
I'm also curious as to why the article mentions credit lines. If SNB policy says it will print unlimited francs to buy euros, why are its credit lines an issue.
I would also note the CFTC reported net shorts in Swiss franc fell 33% last week.
SNB credit lines matter because you need them to trade w/the SNB. If you're a counterparty that doesn't have lines to the SNB, you might end up trading with another such cpty at levels below 1.20 and there is nothing the SNB will be able to do about it, which could lead to actual prints below 1.20. This is what occurred recently. Obviously, as long as the SNB is on the bid at 1.20, this is likely to be very short-lived.Quote from ralph00:
I'm also curious as to why the article mentions credit lines. If SNB policy says it will print unlimited francs to buy euros, why are its credit lines an issue.
Quote from Daal:
The trading books talk about fixed % in order to size your positions. Like 1% or so. Seems totally wrong to me. They are not adjusting by trade quality and EVEN IF THEY ARE, the ratios of adjustment(say 2% for avg trade and 4% for a good trade) is not based on anything mathematically rigorous, why the 100% jump?Why not 150%?
With Optimal F you can get numbers that are based on things that are actually sound and mathematically proven
Not really true, all risk-taking preferences are based on the bettors utility function, and there is no 'mathematical proof' of that, nor can there ever be, since it's dependent on purely subjective emotional responses. There is no formula that can prove your risk preference should follow one slope or another, or even a slope at all. It's 100% rational to be willing to bet 1% on evens odds and 20% on evens + a tiny amount, if that is in fact what you feel like doing.
There's also the pragmatic limitation that trade odds are almost always very vague, it is virtually impossible to calculate them with precision.
Trying to seek objectivity and precision in risk tolerance is futile. All you can say is that risking X% under trade-odds Y can be expected to lead to a drawdown of Z% of capital over a given time period, within a given set of statistical confidence intervals - IF your assessment of the trade odds and the statistical distribution of results is correct (which it almost never is). That is why it's suspect to follow formulas like Kelly, Optimal F, or to make analogies with fixed-odds gambling games or other closed systems that have only small superficial similarities with markets.
Risk taking is something where it's better to be roughly right than precisely wrong, and where robustness is about 1000 times more important than precision. That's why most experienced traders recommend betting conservatively and/or pursuing long gamma strategies where the downside is usually somewhat quantifiable, and where moderate errors in risk assessment don't lead to enormous unforeseen losses. The cost of lack of precision in how much to optimally bet, is relatively trivial compared to the benefit of using a robust approach that is always prepared for the worst. So, best to focus time & effort on the latter.
China just widened their band against dollar. All of the sudden the strength of the HKD on friday is explained. I was scratching my head trying to understand why it was rallying with stocks falling apart. It makes me sick to see all this trading on inside information
I'm currently thinking on how to profit from situations like this. When you see a price move that is not explained and it goes against what the market would normally do. This is the market tipping off there might be insider trading going on and betting in that direction could be profitable
I'm currently thinking on how to profit from situations like this. When you see a price move that is not explained and it goes against what the market would normally do. This is the market tipping off there might be insider trading going on and betting in that direction could be profitable
Quote from Ghost of Cutten:
Not really true, all risk-taking preferences are based on the bettors utility function, and there is no 'mathematical proof' of that, nor can there ever be, since it's dependent on purely subjective emotional responses. There is no formula that can prove your risk preference should follow one slope or another, or even a slope at all. It's 100% rational to be willing to bet 1% on evens odds and 20% on evens + a tiny amount, if that is in fact what you feel like doing.
There's also the pragmatic limitation that trade odds are almost always very vague, it is virtually impossible to calculate them with precision.
Trying to seek objectivity and precision in risk tolerance is futile. All you can say is that risking X% under trade-odds Y can be expected to lead to a drawdown of Z% of capital over a given time period, within a given set of statistical confidence intervals - IF your assessment of the trade odds and the statistical distribution of results is correct (which it almost never is). That is why it's suspect to follow formulas like Kelly, Optimal F, or to make analogies with fixed-odds gambling games or other closed systems that have only small superficial similarities with markets.
Risk taking is something where it's better to be roughly right than precisely wrong, and where robustness is about 1000 times more important than precision. That's why most experienced traders recommend betting conservatively and/or pursuing long gamma strategies where the downside is usually somewhat quantifiable, and where moderate errors in risk assessment don't lead to enormous unforeseen losses. The cost of lack of precision in how much to optimally bet, is relatively trivial compared to the benefit of using a robust approach that is always prepared for the worst. So, best to focus time & effort on the latter.
Its irrational to me to be willing to drawdown 20% with a normal system and the same % with an system that is vastly superior.
I'd be willing to go 25, 30% probably even more with a system or trade that is excellent. That is my point, maybe you think differently but I'm in this for the money
Quote from Daal:
I'm currently thinking on how to profit from situations like this. When you see a price move that is not explained and it goes against what the market would normally do. This is the market tipping off there might be insider trading going on and betting in that direction could be profitable
This is why I like the classic trading books. Kovner made this exact point 24 years ago...