The Finance Book Club

Chapter 6 - Fixed Ratio Trading
Came about as a direct result of studying and breaking down the Fixed Fractional method.

Risk and Reward
Proper MM addresses both risk and reward. Optimal f only addressed overall growth, while fixed fractional only addressed the risks, leaving the growth faltering.

Only antimartingales can be considered in Jones’ philosophy of the markets. Trade size should increase with equity and vice versa. This is the starting point.

First Jones looked at the pros/cons of the Fixed Fractional method.
Pros:

1. Geometric growth is possible
2. Risk is managed with lower percentages

Cons:

1. Higher percentages result in catastrophic risk
2. Lower percentages are inefficient
3. Middling percentages do not balance risk/reward satisfactorily

The fault is ‘unequal achievement’. It’s illogical to demand more profits at the start and fewer as equity increases. It should, if anything, be inverted.

The key is the relationship between the number of contracts traded to the amount of profits required to add an additional contract.

If MM requires $10k to increase from 1 to 2 contracts, then it should require $20k additional profits when going from 2 to 3 contracts. There is a fixed ratio of contracts to required profits. That’s how it got its name.

The only variable in FR is called ‘delta’. It measures the money needed to change lots.

In FR the amount needed for the next increase in contracts increases proportionally which reduces risk below FF; at the same time geometric growth is faster.

Since the risk is lower, a smaller fixed ratio can be used.

Smaller fixed ratio progression example.
The delta of the FR method is only $5000. It therefore only takes $20k to reach the $60k level, whereas FF needs $40k to reach $70k level. (doesn’t it need $60k?)

Formula:
Previous required equity + (No. of contracts x delta) = next level

Suppose:
Starting balance = $10k (first required level)
No of contracts: 1
Delta = $5k

So,
$10k + (1 x $5k) = $15k to increase to 2 contracts

When the account goes above $15k:
$15k + (2x $5k) = $25k
$25k + (3x $5k) = $40k
$40k + (4x $5k) = $60k etc.

Delta means change. It’s the only value to change to fit a particular method or trading style.
The money required per number of contracts is a 1:1 ratio. Multiply the number of contracts and the dollar amount to get the next contract is multiplied by the same number.

If the ratio is 1:$5k then you know going from 10 to 11 contracts will need $50k in profits.

1 x 10 = 10
$5k x 10 = $50k This number is not the same as the required account balance. It is the additional profit needed to get to the next level.

The fixed relationship allows a trader to make other calculations with ease too. For example calculating how many trades it will take to get to the next level.

Suppose a strategy produces $50k in profits over 100 trades.
The average trade is $500.
Since the dollars needed to level up is proportional to the number of contracts we can simply divide the delta by the average trade size to know how many trades it takes to level up.
5000 / 500 = 10. So every 10 trades (on average) this system will increase by 1 contract.

We therefore know that after 100 trades this system will be trading 10 contracts, approximately.
And we also know at that point the total profits should be close to $225k

$190k + (9x $5k) = $235k - $10 = $225k

This is just an estimate though as asymmetrical leverage can and most likely will mess with the results. Jones says a conservative estimate is 90% of estimated profits.

Jones goes on to do mathy things that show the DD using this method is smaller than what we’d expect trading FF. Even though FF produces larger profits faster on the back end, he claims no traders want to use it because of the fearsome DDs it produces.

Jones spends quite a bit of time examining different delta values and comparing them to FF in order to convince us that his method is superior.
“The trader who cannot tolerate the drawdown will not be able to see it through to higher profits. The strategy will be dumped and replaced with another only to go into another drawdown. This is the cycle of most traders. You must take into consideration both the risk and the rewards of any money management method.”

Another relationship that exists in FR is that of the DD to the Delta. If the Delta is $5k and the expected DD is $10k the ratio of Delta to DD is 1:2. In this way you can always know how many contracts you will be trading at the end of a DD of any particular size. “If I am trading 10 contracts with a $2,500 delta and suffer a $10,000 drawdown, I will not drop below trading 6 contracts at the end of the drawdown.”
$10k DD / $2500 delta = 4 ‘delta levels’ or contracts.
10 - 4 = 6 contracts after the DD

This formula will you each contract level change without the need for a table.
[(# of contracts2 - #of contracts) / 2 x delta = minimum profit level

Suppose:

1. Trading 10 contracts
2. Delta $5k
3. 10 x 10 = 100
4. 100 - 10 = 90
5. 90/2 = 45
6. 45 x $5k = $225,000 = minimum profit level; this is the level at which the contract increase or decrease takes place, from 9 to 10 contracts or from 10 to 9 contracts depending on whether profits go above below this number.

By changing the above formula to use ‘+ # of contracts’ instead of minus:
[(# of contracts2 + #of contracts) / 2 x delta = maximum profit level

1. 10x10=100
2. 100+10=110
3. 110/2 = 55
4. 55x$5k = $275k, the upper profit level for trading 10 contracts. This is the upper level for 9 contracts and the lower level for 11 contracts. “Since I am able to calculate these levels as well as calculate the maximum levels that any drawdown will decrease the account, I know the exact dollar risk at any given time.”

Jones carries on with more examples of this sort, but it seems fairly irrelevant in this computer era where the machine keeps track of everything. Or maybe I’m just succumbing to the ‘quick fix’ nature of our culture??

Applying the Fixed Ratio Method to Stock Trading
Stock trading usually has only a 50% margin rate, whereas in commodities it can be as low as 10%.

Stock trading allows odd lots. (but what does this have to do with anything?)

Jones devotes pages of arcane tables to explain the miniscule differences between stock and commodity trading margins, but in the end he says all we need to do is to start out with an extra $500 in the account to nullify margin problems while using the same money management concepts. Padding!

Trading a basket of stocks? Just make the same margin allowance for each stock.

5 x $25 (margin for average 1 share) = $125
Trading 100 lots of each would require a margin of $12,500.

Chapter 7 - Rate of Decrease
Usually most systems increase risk at the same rate that they decrease it.

Jones proposes an asymmetric method. The rates of increase will not necessarily be the same as the rates for decreasing risk.

The function is to protect profits and enhance geometric growth. You could also call it ‘asymmetric leverage’ abandonment.

You cannot have your cake and eat it too with this strategy; you will have to choose either to protect profits or enhance growth but not both.

Protecting Profits
“Decreasing risk faster than it was increased will protect profits during drawdowns.”

It will also allow more aggressive reinvestment during good times. The primary reason to not be aggressive when reinvesting is fear of drawdowns, but if risk is decreased faster during drawdowns, the DDs will be smaller.

This has at least one drawback

1. It increases the effects of asymmetrical leverage. The ability to gain back losses decreases as you decrease risk. Because profits will be smaller initially.

The formula for finding a new decrease rate:
CL = Current Level
PL = Previous level
X% = variable percentage (a new variable besides delta!)
CL - [(CL - PL) x X%] = Next level of decrease

Suppose CL = $275k and PL = $225k
$275k -[(275k - $225k) x 50%] = $250k, the new level of decrease. Instead of decreasing contracts when profits decrease to $225k, they are decreased $25k earlier.

Does anyone understand what these tables are supposed to show? The math doesn’t even add up. Seriously, if anyone can explain this to me, please, show off your brilliance.

The problem I have with this is not that it’s bad graphic design, bad math, or bad explanation, it’s that if you put in something confusing in your book I have to wonder if the problem is me, or you. If you’re bad at explaining things why am I trying to learn from you? If you’re intentionally trying to confuse me, why should I trust you? These tables are awful.

Jones runs through a complex example that compares what happens if you use the decreasing risk after suffering a DD and then gain back on a series of trades. He concludes that profit protection provides a 57% benefit at a cost of a 7% loss of profits on the upside.

He then introduces the idea of the ‘reincrease switchback’ which increases risk faster so that you can return to pre-decrease levels as quickly as possible.


But it gets worse? WTF??

Double WTF

This is starting to get ‘complicated’. My question is, what if your run-down and run-up aren’t smooth and continuous? What if you are in a choppy period and winning a few, losing a few; how would you keep this straight?

Jones addresses this (sort of).
“A few risks are involved that traders should consider if using the switching method. The reason the lost profits are gained back is because the number of contracts increases from five to nine in one jump. This is great if trading continues on a positive run but if the very next trade becomes a loser, it loses with nine contracts, not seven. You then would have to jump back down to four contracts, which would increase the effects of asymmetrical leverage all the more. Use caution when applying this method around the levels at which the rate of reincrease is switched.”

It’s funny how trading book authors always say “use caution” when butting up against an unavoidable and insoluble dilemma caused by their advice. I should write a book about trading and that’s the whole book. USE CAUTION.

Further to this Jones says the switch method should only be used “when there is a significant difference in the account sizes before the positive run begins” but he doesn’t say how much is ‘significant’.

Increasing Geometric Growth by Abandoning Asymmetrical Leverage
Hopefully you get the idea behind asymmetrical leverage, because Jones re-explains it using coin-flipping examples and I don’t feel like going back there.

The point is that decreasing risk protects profits, permitting geometric growth, more than it hurts profits due to asymmetric leverage.

Somewhere in Between
This is about using some method of asymmetric risk that is less aggressive than the above.

“A trader may want to stay at the highest number of possible contracts until the drawdown exceeds the largest expected level and then decrease. By applying this method, the trader is waiting to bail out until the last possible minute. This also allows the trader to avoid any asymmetrical leverage for all drawdowns that are smaller than the largest expected drawdown. Then, if that drawdown is exceeded, the trader will protect profits from that point on until the drawdown is over.”

Another tactic is to decrease by ½ the speed that contracts are added.

There is no limit to the variations that can be attempted, says Jones, but he advises sticking with mathematical relationships rather than using trade performance relationships - like counting the number of consecutive losers, etc.

Chapter 8 - Portfolios
This age-old concept is all about diversification to avoid risk.

If you can afford to build a portfolio but instead choose to play a single instrument or option you are gambling, says Jones.

However Jones then reverts back to coin-flipping examples to prove his point and I feel that this neither teaches us anything nor is it particularly applicable to the real demands of trading.
So - not going to summarize it.

If you believe in portfolio theory, then you’re sold. If you don’t, this chapter won’t sell you.

The second part of the chapter is Jones arguing that Fixed Ratio is better than no MM, but obviously we’ve accepted this argument else, why bother?

Chapter 9 - Market Weighting
This chapter addresses a common grievance that Jones encounters in his seminars. Traders want to ‘equalize’ contracts between different markets, but Jones argues this is unnecessary and counterproductive.

For example if you trade S&P and corn, and the corn system has a DD of $5k and the S&P system has a DD of $15k, the temptation is to trade 3x as many corn contracts as S&P because then the DDs will be ‘equalized’. Jones says no.

He says traders do this only because it ‘sounds’ logical.

His proof consists of the following facts.
Trading the combined systems produces a DD of $12k.
The reason it’s $12k and not $20k is that the largest DD doesn’t happen ‘at the same time’ (but why couldn’t it?).
If you trade 3 corn contracts, any DD in corn will happen at the same time in all 3 contracts, thus producing a guaranteed $15k DD. This negates the diversification benefit you get from trading two markets.

Facts about the S&P system:


Facts about the corn system:


Facts about the combined system:


Now suppose we ‘equalize’ by adding 2 more contracts of corn for a total of 3 corn contracts. The results are:


“By adding two additional corn contracts, the drawdown increased by at least a full contract. Therefore, we lost the benefit of noncorrelating drawdowns of one of those contracts.”
“According to drawdown and the fact that money management is more efficient with lower drawdowns, the logical thing to do is to trade one corn contract with one S&P contract.”

It’s better to add a new market than to add additional contracts to an existing market.
You increase the chances of net profit and non-correlating DDs.

Suppose we add a bonds system.


The results of combining performances of all three markets:

Compare this to the performance from trading 1 S&P and 3 corn. Net profit is larger. DD is smaller.

What I am reminded of is Perry Kaufman’s advice about trading unequal amounts of different instruments in a portfolio. The only reason to do so, he said, is if you think that there is a greater probability for one of the trades to work out over the other, in which case it makes more sense to just trade more volume of the higher probability trade. His philosophy was that balancing things out was important since we can’t know how volatility will affect any given instrument in the future. Was he running afoul of Jones' argument that this increases correlated DD risk? Or is Jones correct that balancing volumes maximizes diversification in general (not just coincident DDs)?

The answer is far from simple to work out, and it’s further complicated by the differences in trading systems employed by the two men, and any translation of those systems that we would employ. This is why even very smart people like Jones and Kaufman, trade differently, and why the marketplace is not easily dominated by any single approach, no matter how sophisticated.

Further, I’m suspicious of any theoretical approach that is followed too slavishly. The only way to be sure is to try things out in real life and that’s slow and dangerous.

Chapter 10 - Market Weighting Through MM, Not Before It
Jones tries to head off any arguments about which is the superior weighting method by taking a different approach.

Suppose we are trading 3 different markets. S&P, Bonds, and corn.
Largest DDs:

1. Corn: $8k
2. S&P: $12k
3. Bonds: $4k

Instead of tinkering with contract equalization, market-weighting through MM applies a different delta to each market.
However it generates the profits to reach each new level from the combined profits of all 3 markets. Now this seems clever.

Suppose that the combined DD of all three markets is $12k. Then the original MM method would increase contracts for all three markets with a $6k delta. However 75% of the combined DD is attributable to S&P, while bonds and corn is only 25%.
So, adjust the delta.

1. Corn: $2k
2. Bonds: $4k
3. S&P: $6k

All three markets start with a single contract, but corn will increase the fastest or first. This is not the same as ‘equal-weighting’ before MM, which is what we saw in the last chapter. This method applies different weights according to degree of risk.

“If one market offers a much smaller degree of risk, we aren’t increasing that risk to meet the degree of risk of the other markets; rather, we are allowing the market to increase contracts more efficiently than the markets with greater risk.”

This is more efficient. It also allows delta to be larger than when trading a solo market. You could use ¾ instead of ½ delta. This can yield more profits but keep DD stable.

However, Jones ‘tests’ this idea by giving us a fictitious trading history with no losing trades. WTF? How is this supposed to convince anyone? Obviously testing this within the confines of a book is not going to be easy, however. Needs computers.

Chapter 11 - Other Profit Protecting Measures
Here Jones addresses topics covered in the first chapter in more detail.

Consecutive winners/losers. This idea comes from the gambling arena and has no mathematical basis.
I believe he’s right both from anecdotal experience and from a statistical basis. No need to go into it much.

Jones goes on a lengthy rant against gambling and rigged games.
I agree with everything. Why would you get into a ring with someone who is going to knock your block off? Then again, why trade against traders who have many multiples of millions of dollars against your puny thousandaire account? The only reason is because you think there is a way to find an edge.

Jones goes on an even lengthier examination of the statistical nature of streaks and dependence.
Now, even though I think he is fairly adept mathematically, here he is completely in over his head. He admits there is statistical dependence in price series. Then he gets somewhat …opaque.

“The only way some type of dependence can exist in trade sequences is if the dependency in the markets is somehow transferred to the trades that are being taken. This is no easy task to accomplish.” I’m not even sure what he means here or if this is a dismissal of trend-trading?

Trading the average of the equity curve. I agree, it’s silly.

“The number one reason for business failure is undercapitalization. I would also deem that it is the number one cause of trading failure.” So my $50 challenge is probably doomed? Probably.

The rest of the methods listed are possibly helpful in reducing the risk of complete ruination (worst case scenarios), but Jones says they are not proven mathematically to improve performance.

Average Equity Curve Trading With a Negative Expectation
No MM method can change negative expectancy to positive. However that doesn’t mean it won’t sometimes happen due to random luck.

Two Consecutive Closes Below the MA (of the equity curve)
“This method requires that the equity curve move below the moving average but also requires a confirmation by having the following equity point close below the moving average.”

Dollar DD with 30% Retracement
“Instead of using a moving average to indicate when to stop taking trades, simply use a set dollar amount for the drawdown to exceed. After the drawdown exceeds this level and trades have ceased, require that the dollar size of the drawdown be retracted by 30 percent before starting to take trades again.”
If the largest DD is $8k, stop taking trades once it exceeds $9k. If it goes to $12k and comes back up by 30%, meaning the drawdown would have to decrease from $12000 to $8400, start tracking trades again.

Trend lines using the equity curve.
“Using trend lines on the equity curve is another way of cutting the larger losing streaks of any method or system down to size. Trend lines can be used with the equity curve by drawing a line between the two most recent low points of the equity and extending it into the future. If the equity breaks the line, trading is halted.” How does equity recover if we aren’t taking any more trades though? Imaginary trades?

Chapter 12 - Risk of Ruin
Jones begins this chapter very controversially dismissing the whole notion of RoR as nothing more than a useless statistic.

He goes through a ‘Zeno’s paradox’ type analysis (again using coin flipping) of why RoR is of no use to traders and concludes that the only thing to be done is to have an account big enough to handle unavoidable risk. He does address the fact that RoR could be used to find out where the barrier is between ‘too small’ and not ‘too small’ but again defaults to the common wisdom of using ‘common sense and logic’ when trading with a small account.

“My suggestion for small accounts would be to stick to trades that, in and of themselves, have a high probability of making money.” You don’t say?!

He points out events like platinum going below the price of gold in January 1997 (advising a pairs trade).

However if this is the right way to trade, why is it only for small accounts? I’m beginning to detect a soupçon of chicanery.

Chapter 13 - The System
The purpose of this chapter seems to be simply to rant about the current state of affairs in trading. There is an undue emphasis on systems and methods and not enough on MM! Well, sure.

Long-term success can be attained with simple, logical, mediocre systems and methods (but also using proper MM, implied, surely? Or something. I can tell you after years of reading FF that he's not wrong.).

On the other hand, besides being a simple padding chapter, I wonder if I detect some misdirection? If you don’t have a working system, but just one variation on an existing MM theme, every problem looks like it has one solution - yours. And don’t forget to buy the Performance 1 software and books!

Jones goes on to describe statistics he likes to look at when evaluating a trading system.

1. Total Net Profit
   1. Gross profits minus gross losses
   2. The broadest view of what the system can do
   3. The time frame is crucial to consider

2. Max DD
   1. DD is the distance between a high point in equity followed by a lowest point in equity until a new high is made.
   2. The DD lasts until the previous equity high is exceeded
   3. This number can be misleading as there may be multiple smaller DDs that are just as bad or worse, or the max DD may be so rare that it’s not as bad as it seems.
   4. There is no guarantee this ‘max’ DD won’t be exceeded in the future
   5. Good for making an educated guess on what to expect regarding risk
   6. Evaluate this with the average DD compared to average yearly profit

3. Mathematical Outcome (Expectation)
   1. Covered in chapter 2
   2. [1 + (average win / average loss)] x Winning % - 1
   3. Jones prefers this be over 0.6
   4. Can’t be used to gauge future probabilities but only to examine the strength of an historical record.

4. Average Trade
   1. The take-home statistic
   2. Total net profit / total number of trades
   3. A higher average trade gives you more room for error
   4. Jones won’t consider a number < $250

5. Average Win/Loss and % Profitable
   1. Not valuable alone, but together they rank on top
   2. Gauges room for error

This is actually quite useful to an old algo trader like me. One of the hardest things in trading is trying to decide if a system is worth going live and if it's an improvement over an existing one. It's a bit like drafting a new player and trying to evaluate how they'll perform in the big leagues.

If a system or method is 50% correct with a win/loss average of 1.00 it is break-even.
So if winning trades are 50% the win/loss must be over 1.00.

The chart above shows the break-even points for different win% and W/L ratios. His rule of thumb is that a ‘very good’ system should be 1.0 more than the break-even value for the win% reduced by 10%.

So if you have a method with 70% winning trades, reduce that to 60%. B/E is around 0.70 so the system’s ratio would need to be 1.0+0.70 = 1.70 with a 70% winning strategy.

If the strategy had a 50% success rate, win/loss needs to be 2.50. This is ‘about as close to a holy grail as you are going to get.’ But without knowing the average trade, this would include ‘eating like a bird and shitting like an elephant’ systems, no?

 

1. Average DD
   1. Not the same as largest DD
   2. Take all DDs and average them
   3. Filter out any DD that is <= 3*avg losing trade
   4. Compare with largest DD
   5. Largest DD/Average DD should be a 2:1 ratio; any less than this means the largest DD is likely to be much larger (and coming up)

2. Average DD / Average yearly profit
   1. Lets you know what to consistently expect
   2. Ideal is 1:2 ratio with at least 1:1 acceptable.

3. Ratio of Largest Win to Average Win
   1. Also largest loss / average loss
   2. The value of this ranks near the bottom
   3. Used to gauge how likely another largest win will be
   4. If it largest win is greater than 3-4x average win don’t expect to see it again soon
   5. Losses greater than 3-4x size of average loss are likely caused by slippage and gaps; should be rare; if the ratio is less than 3, these might be the rule not the exception

4. Profit Factor
   1. Gross profit / gross loss
   2. If other stats aren’t conclusive require this to be at least 2.0
   3. This means you win 2x as much as you lose.

I take it back. This is not a padded chapter at all. This is golden info.
“I like to pick several that cover both risks and rewards and a few comparisons between the two. If everything lines up on those, most everything will line up on other statistics as well. I have done my homework. No matter how many statistics you look at and no matter how many fall within your designated requirements, it will not change how the system is going to perform.”

“I do not like to use statistics to gauge how much money I am going to make by trading any particular system. I use statistics to gauge when I should say “uncle” and when I should stick things out.”

Don’t trust black boxes. They can look great due to curve-fitting.

“Although other logics may work, cutting losses short and allowing profits to ride is king in a trading method. It is the most reliable logic that can be applied across the board.”

Forget about systems based on mumbo-jumbo (astrology) and prefer simple over complex.

Rules for filtering out losing systems and methods (not failsafe)

1. Make sure the body of the logic is related to what you are trying to accomplish.
2. Make sure the statistics are in place and meet the minimum requirements.
3. Make sure the logic of the entries and exits relate to the goal you are trying to accomplish.
4. Make sure the method is as simple as possible to capture the logic.

A Simple Trading Method
Oh boy! I wasn’t expecting this; were you expecting this?
About as simple and logical as they come, says Jones.

 

1. Trend-following method
2. No exit rules aside from protective stops.
3. Reverse on signal change. Exit on profitable trade or stop out.

Buys:
1. Requires the average close of X days to be greater than the same average close Y days ago.
2. Requires the close to be less than the close Y days ago.
3. Requires the close to be greater than the close Y + X days ago

If these conditions are met - buy on the open the next day.

Sells:
1. Requires the average close of X days to be less than the same average close Y days ago.
2. Requires the close to be greater than the close Y days ago.
3. Requires the close to be less than the close Y + X days ago.

If these conditions are met - sell on the open the next day.

Defaults:
X = 20;
Y = 3;

Here the average close of the previous 20 days must be greater (for longs) than the average close of the last 20 days, starting 3 days ago. This means the average close of the last 20 days is sloping up. Today’s closing price must be less than the closing price 3 days ago. This helps find a pullback before entry. Finally, it requires that the close, even though it is less than the close 3 days ago, is greater than the close 23 days ago. This verifies the up-slope MA.

One very glaring omission - where does Jones place his stops to get these (unbelievable?) results?! LOL. It’s like giving us a treasure map but neglecting to mark the X.

The system itself is a classic ‘buy the dip’ system that should work. However can it produce these fat profit factor results? I guess we’d have to build it and test it to find out. And so that's what I'm going to do. I've actually been working on an EA to do exactly this for the past few days, trying to debug as much as I can. I will reveal it in the fullness of time.

I will say my BS radar is on high alert, especially since we seem to be missing vital details about entries, exits and stops. ‘Simple and logical’ it may be, but possibly too simple. If it goes well, credit the system, if it fails, the fault is mine.

Here is the EA I promised. I called it Jonestown as it's based on the system by Ryan Jones and it could massacre your account if you're not careful.

My code writing abilities have improved since I started learning a few years ago but I'm still a rank novice. ChatGPT was helpful but isn't a substitute for human mentorship yet, so if any expert coders follow this thread and want to make improvements, please feel free. Suggestions are also welcomed but may be rejected for any reason.

It attempts to follow the Ryan Jones system as he describes it in his book as closely as possible. However it uses simple moving averages to calculate the entry points. I'm unsure if this is exactly the same as what Jones had in mind. It should be, since an SMA is just the average of a certain number of bars of closing prices.

The 'protect profits' mode was a real headscratcher, especially as Jones does not describe it clearly in my opinion. I did not have any luck with it, and I'm not sure if it's on me and my inadequate bugfixing, but I left it in, disabled, for you to play with. Maybe you can find a setting or a bug that I missed. Maybe you can use this as the basis for some holy grail or discover that this is the holy grail, just misconfigured.

I had good results testing on AUDUSD, but very poor results on the Dow Jones, which is not intuitive, though both instruments can trend for long periods.

Chapter 14 - Optimization
Seems out of place in a book about MM, Jones admits (well, he lampshades it, actually).
In fact it’s a classic trick. Jones is leading us down a primrose path where he optimizes a system and then points out with flawless logic that it’s been curve-fit and the future results won’t match, making the whole process USELESS. However he claims there is value in the optimization process. Whaaaa?

The Optimization Process
“For example, the optimization performed on the Swiss franc market with the simple moving average crossover system went through 496 different parameter tests. Each one of those tests produced a different set of statistics. To be able to draw some practical conclusions on what to expect from any given trading system, it is much more beneficial to know what kind of numbers the bulk of those tests produced than what kind of numbers the single best test produced.”
Monte Carlo method?

Jones compares statistics from one in-sample test then does an out-of-sample test.
“What we want to do is trade a system where the odds of making money are in our favour regardless of which combination or set of parameters we choose.”

He reaches the conclusion that the number of possible profitable parameters is too large to possibly know what will be a good value in the future, but for some systems the number of profitable vs. unprofitable parameters means that the odds of picking values that are profitable outweighs the odds of picking a turkey. He also completely dismisses the value of hypothetical testing results. However, just a chapter or so ago he thought it was OK to use made-up trading results to prove his portfolio-weighting theory? What gives?

Chapter 15 - Commodity Trading Advisors (CTAs) and MM
A chapter for CTAs to broaden their options and to give investors looking to invest in a CTA something to look for.

CTA/CPO are fund managers. Their accounts range from 100,000 - 999M dollars.

Large CTAs manage money very conservatively. Even an 8% DD could cause a mass exodus from the fund.

They generally use FF MM and risk only a fraction of a percent on each trade. This sounds paltry but with a $50M fund and a single trade with a $3k stop for every $600k under management that’s 83 contracts.

Just like traders, most CTAs focus on entries/exits, not MM.

Jones advises:

1. Switch to FR from FF (naturally)
2. Divide the funds into parts so that you can diversify systems and methods

Jones goes so far as to suggest dividing the fund into 12-15 parts. The resulting smaller segments trade smaller numbers of contracts using a diversified portfolio of systems.

He spends a lot more ink convincing us that the FF is inferior to FR but we conceded this point long ago. Why belabour it? He points out that the chances of all 15 systems going into the largest DD simultaneously is nearly impossible and it would mean the systems designers were negligent.

Smaller CTAs with their smaller accounts necessarily take on more risk. However they can also see more growth. They can risk as much as 2% on any given trade in a bid for higher returns.

Jones mentions John Zervas who is completely anonymous on the internet to this day. If he became a big time adviser it was awfully hush hush.

“I have never managed money for a fund, nor do I have any desire to do so.” No, it’s much more profitable to sell books and software.

Chapter 16 - MM Marriage
Although FF is superior to FR in both risk management and reward growth, when an account reaches a certain size, one of the drawbacks of FF can actually help one of the drawbacks in FR.

The drawback in FF, as you’ll recall, is low reward compared to total risk when keeping risk relatively low. To keep risk low, a low risk percentage is needed. To keep risk at 10% if a $10k DD is sustained, you’d need one contract per $100k. That one contract must produce another $100k before another contract can be added.
So the two problems are:

1. If we assume that trades are independent then the $10k DD can actually grow to be more than that if following trades also lose.
2. The second problem is growth potential is negligible.

The first problem will never go away with FF. The second problem decreases as the contracts traded gets larger. So it can go away over time.
In contrast FR works almost the opposite way. At the start it’s growth potential is faster. It gets slower as the account enlarges.

“Fixed Fractional method = Increase in growth rate with constant risk
Fixed Ratio method = Constant growth rate with decreasing risk”

As he is wont to do, Jones devotes pages of mathematical equations to show this is true. Stipulated.

So, it makes sense to switch from FR to FF at some point. What point? When the growth rate of the two methods cross. You can also use the risk percentage.

Jones provides this somewhat rickety chart.

“If you wanted to use the Fixed Ratio method until the risk was lowered to a constant 12 percent should a $10,000 drawdown per contract be sustained, then the switch would not occur until the Fixed Ratio level of increase crossed that risk level of the Fixed Fractional increase. This means that the Fixed Fractional method would increase one contract for every $83,333. The Fixed Ratio method would increase one contract for every $5,000 as before but would reach the risk level of no more than 12 percent at 29 to 30 contracts. This is the level at which the account balances cross in Figure 16.1.”

The vertical axis is the FF calculation that trades one contract per $83,333. Low risk from the start, but ‘impossible to implement’. The horizontal axis represents the FR method using a $5k delta. Around contract #17 the methods cross, and that’s where you should switch to FF. Clear as mud?

Chapter 17 - Putting it All Together
Having goals is pointless without a plan.

You have to stop trading before you form your plan.

 

1. Take into account what you have done
   1. Write down the methods, systems, strategies, and trades you have ever made.
   2. This does not mean list every single trade you have ever taken.
   3. However, if you have ever opened an account for the sole purpose of writing options, summarize the experience.
   4. Write down as much detail as possible including
      1. what events lead you to start trading a particular method,
      2. the account size you started with,
      3. why you started with that account size,
      4. the type of trading,
      5. the frequency of trading,
      6. whether you enjoyed taking the trades (or hated it).
      7. Write down how long you traded the method,
      8. whether you stuck with it, (if not, why),
      9. the size of drawdowns as well as drawups.
      10. Don’t forget to include time variances. How long did the drawdowns last compared with the drawups?
      11. How long did you stick with the method?
      12. Finally, write down when you quit trading the method, what the final outcome of the experience was (gain/loss) and what you did with the account afterward.

2. Write Down Your Goals
   1. Goals begin with how much capital you are starting with.
   2. When you have a goal of making $1 million in the next five years, is that goal to reach $1 million risking the entire starting account to achieve it, or do you only want to risk 50 percent of the account to achieve the goal? The risk tolerance becomes part of the goal.
   3. The goal may also include other things such as the manner in which you want to achieve the end dollar figure.
      1. Do you want to achieve it by spending four hours a day researching the markets after you get home from eight hours of work?
      2. Do you want to be able to achieve the goal and be completely hands off for time to spend with your family?
      3. Is part of your goal to be able to quit your current job within a certain time period?

   4. Also include within the goals the sacrifices you are willing to make.
      1. No amount of wealth is worth sacrificing your wife, husband, or children, but there are sacrifices. Sacrifice the things that have little impact first. Maybe you will have to cut out your three golf games a week or soups at lunch.

3. Develop a Plan of Action
   1. Methods
      1. Develop a diversified group of methods and markets
      2. Don’t overtrade - make sure you are within starting balances as described in chapter 4.
      3. A well-diversified portfolio may consist of one long-term trading method, one short-term trading method and possibly an options trading strategy or some other type of system unrelated to the first two. Be careful not to have too many similar strategies as you will often be on the same side of trades in both systems.
      4. “Easy Money Trading” - look for unusual market events
         1. Jones use the example of Orange Juice in April 1997. OJ was done to $0.68 which is extremely low. He figures OJ can hit $1.30 within two years and it does on Oct. 10, 1998.
         2. He retells the story of Gold/Platinum.
         3. He recommends John Zervas ‘Smart Trading Market Letter”. I can’t find any trace of it anywhere. The phone number listed does still have ties to Zervas though.

   2. Money Management
      1. This book already delivers the goods, says Jones
      2. Gauge the MM strategies against your goals
      3. Jones takes the opportunity to plug his software one more time
      4. It is best to understand something before you do it otherwise you might not follow it.

4. Concentrate on your Strengths; Delegate your Weaknesses
   1. Dan Marino (American Football quarterback) doesn’t kick field goals
   2. You should delegate what you aren’t good at (algos!)
   3. This leaves more time for what you are good at

5. Prepare a Backup Plan
   1. The only traders who fail are traders who quit.
   2. You should always have room to continue trading if your first plan produces losses.
   3. If you risk your whole account on one plan don’t have a backup plan.
   4. Worst case scenario should only lose 40% of the portfolio.
   5. The backup plan should be more conservative than the original failed plan. For example, only trading ‘easy money’ trades.

6. Prepare for Additional Strategies and Markets
   1. Adding strategies and markets is part of proper MM
   2. A good rule of thumb is to add to the portfolio once every 6 to 8 contract increase in the current strategies.
   3. It is a good idea to explore trading strategies that are unrelated to what the plan is currently trading. (why I’m here)
      1. Longer-term trend following
      2. Short-term swing trading
      3. Look at breakout systems or the same systems in different markets

7. Optimization Stats and Portfolios
   1. This just seems to be a repeat of what he did in Chapter 14. Just a complete re-roll of the same thing except applied to the Swiss Franc this time. Padding.

“ What is relevant for one trader will not be for another. The basics behind these principles are applicable for every trader. How exactly to apply them to your own trading is a question you alone can answer.”

Jones finishes the book with advice for those who are compelled to trade intraday. Go ahead and open a separate $10k account, blow the money but then stay focused on what really works. “Do not alter the plan. Do not pollute the plan with hunches, trading tests, or other things that you have not tested or thoroughly researched. Keep focused and play to win.”

Solid advice at the end. And that's it for another book Next: my impressions.

Hi clemmo,

and then after that book .... Time Compression, maybe ?