Make Money While Cruising - List

[Delfin]

Originally Posted by Sondor

Dennis famously turned an initial $5k into over $100 million. He's obviously fully on the risk-taking bandwagon, but even he has a level of risk tolerance he's not willing to exceed, and that level of risk (25%) is FAR less than Moore's system as touted!

No one is touting anything. I simply explained what my real world experience has been and that has been confirmed by Boden. I'm not a broker, adviser, or a professional in this area at all. Certeza apparently is, and makes money managing other's money. IMO, his analysis is wrong, but what do I know? If he or someone else has another investment strategy that delivers a near 100% return per year over time, they should pursue it. No argument from me on that score.

Moore's method is simple. They identify historic spreads that have rational reasons for being profitable. Then they model 15 years of retrospective returns to identify an entry point and exit point that has a minimum of 87% likelihood of being profitable. On that basis, they make their recommendation. Since the past is no guarantee of the future, the actual number of wins and losses is lower, but year in and year out extremely profitable. One of the reasons it is profitable is because their recommendations span all available commodity categories where significant liquidity exists. That just means that they aren't recommending Oat spreads or Milk spreads because those markets are too thinly traded. Some years currencies suck, but metals make up for it. Other years metals suck but currencies make up for it. That's the whole point and the reason why I really don't get the whole "risky business" meme. If it was so damn risky, why has it been profitable 26 out of 27 of the past years?

As a practical case in point, I'm looking at my portfolio screen for an account in Interactive Brokers that trades a single spread set. The total unrealized profit against 18 active spreads is $14,082. 6 are currently underwater with unrealized losses of $1,264.00, meaning the remaining 12 have unrealized profit of $15,346. I think I'll go out on the boat today.

[Certeza]

Originally Posted by Sondor

I've been enjoying this thread and greatly appreciate Certeza's real-life risk expectancy examples, but I'm a bit baffled. We've all been reading the same posts I assume?

Why is it so hard to understand that Certeza is simply demonstrating that risk should be demonstrated in an apples to apples basis. Not for the SEC, not for blowing smoke etc... but to keep trading strategies truly comparable?

.................................................. ........................


Certeza is simply clarifying that risk needs to be fairly evaluated for any investment, and expected returns are based on the level of risk you are willing to take.

Taking Slick Wily risks for Farmer Bob level returns is not a great investment strategy.

If you are going to take Slick Willy risks, at least earn Slick Willy returns.

.................................................. .................


There were plenty more lessons, but the key takeaway for the sake of our discussion is number two as again it relates to risk vs reward. Note the cap of a 25% drawdown? Risk is risky and should be rewarded accordingly, but never taken on without fully understanding the... RISK

Dennis famously turned an initial $5k into over $100 million. He's obviously fully on the risk-taking bandwagon, but even he has a level of risk tolerance he's not willing to exceed, and that level of risk (25%) is FAR less than Moore's system as touted!

(several books have been written on or even by the Turtles if you're interested - great reads!)

Thanks, Sondor, for confirming that at least a couple people might've been paying attention to the discussion. I don't know why the concept of risk must be spelled out. The concept of considering reward in terms of risk is not a novel concept. It is not theory or conjecture. As soon as someone posts an equity curve or return chart, everyone can immediately determine how the strategy stacks up against all others. The mandated performance reporting standards ensure that people don't have to decipher the graph, and we can get to an apples to apples comparison very easily. In the real world, how much a person is making is only considered relative to how much they are risking and how often the risk materializes.

The history of the Turtle Traders is a good read for anyone getting into trading. The 25% max drawdown rule of thumb is pretty good for the beginner and intermediate trader, because at that level they don't yet fully understand how to utilize expectancy ratios to determine position sizing for proper risk control. But as explained earlier, once someone has a full understanding of expectancy and how to use it, the rules of thumb aren't needed.

Hypothetical example. When running a scan one day let's assume I find a trade like this;

0.0001% chance of exceeding 4% loss
2.2% probability of a loss >1.7%
18% probability of any loss at all
82% probability of any net profit
75% probability the one month return will be an account gain between 0.2% and 0.8% with an average of 0.45%

From there I decide my acceptable confidence interval. Maybe I'm risk averse and want statistical certainty and I'm thus looking at the 0.0001% chance of a 4% loss and consider that my max loss. OTOH, maybe I'm more aggressive and only care about the 2.2% probability of a possible 1.7% loss. So let's assume that it is the latter and I use the 2.2% chance (97.8% confidence interval or 2 standard deviations), but my tolerance for risk isn't to possibly lose 1.7%, instead I'm willing to risk a loss of about 5%.

I could make the same trade, but increase the trade size to 3X larger (or alternatively maintain the trade size, but find three different trades of this quality), so now my expectations for the trade look like this;

2.2% chance of a 5.1% loss
Still 82% chance of a net profit
Average one month return is now 1.35%
Most likely scenario is a monthly return between 0.6% and 2.4%

Hypothetically, if a trader could find this trade consistently over the long term, statistically he'd expect his performance to look like this;

Largest anticipated drawdown ~7% (happens on average once every 4-5 years)
Approx two negative months per year
Approx annual returns of 17.5%

To extend it a little further, if he was willing to risk 75% drawdowns on trades like this, he'd be looking for at least 357% annual returns. But the kicker is that about once every 4 years he'd lose 75% of his account value, and there is even a small possibility of a complete 100% loss. Even with those outlandish returns, it isn't worth it for most people, so they are wise to stick to max drawdowns of less than 25% like you suggested. In my experience, most people who really understand risk prefer to limit it to less than 15%.

[Certeza]

Delfin,

It's become pretty obvious through your comments at this point that you have very little grasp on the concept of risk or how it applies. If I asked you what your true risk is, based on historical performance, you wouldn't be able to tell me. I mean, you've already demonstrated that you don't even understand the basics of how to calculate performance.

As compensation for going out of my way to explain these very elementary concepts, I get to put up with a bunch of underhanded insults suggesting that nothing I say is valid because the evil professional must have some ulterior motive to steal people's money. It's tiring at this point.

Best of luck in your trading.

If anyone else would like to understand these concepts I'd be happy to explain them further, but I'm tired of bashing my head into the wall.

[cwyckham]

Originally Posted by Certeza

Thanks, Sondor, for confirming that at least a couple people might've been paying attention to the discussion. I don't know why the concept of risk must be spelled out. The concept of considering reward in terms of risk is not a novel concept. It is not theory or conjecture. As soon as someone posts an equity curve or return chart, everyone can immediately determine how the strategy stacks up against all others. The mandated performance reporting standards ensure that people don't have to decipher the graph, and we can get to an apples to apples comparison very easily. In the real world, how much a person is making is only considered relative to how much they are risking and how often the risk materializes.

The history of the Turtle Traders is a good read for anyone getting into trading. The 25% max drawdown rule of thumb is pretty good for the beginner and intermediate trader, because at that level they don't yet fully understand how to utilize expectancy ratios to determine position sizing for proper risk control. But as explained earlier, once someone has a full understanding of expectancy and how to use it, the rules of thumb aren't needed.

Hypothetical example. When running a scan one day let's assume I find a trade like this;

0.0001% chance of exceeding 4% loss
2.2% probability of a loss >1.7%
18% probability of any loss at all
82% probability of any net profit
75% probability the one month return will be an account gain between 0.2% and 0.8% with an average of 0.45%

From there I decide my acceptable confidence interval. Maybe I'm risk averse and want statistical certainty and I'm thus looking at the 0.0001% chance of a 4% loss and consider that my max loss. OTOH, maybe I'm more aggressive and only care about the 2.2% probability of a possible 1.7% loss. So let's assume that it is the latter and I use the 2.2% chance (97.8% confidence interval or 2 standard deviations), but my tolerance for risk isn't to possibly lose 1.7%, instead I'm willing to risk a loss of about 5%.

I could make the same trade, but increase the trade size to 3X larger (or alternatively maintain the trade size, but find three different trades of this quality), so now my expectations for the trade look like this;

2.2% chance of a 5.1% loss
Still 82% chance of a net profit
Average one month return is now 1.35%
Most likely scenario is a monthly return between 0.6% and 2.4%

Hypothetically, if a trader could find this trade consistently over the long term, statistically he'd expect his performance to look like this;

Largest anticipated drawdown ~7% (happens on average once every 4-5 years)
Approx two negative months per year
Approx annual returns of 17.5%

To extend it a little further, if he was willing to risk 75% drawdowns on trades like this, he'd be looking for at least 357% annual returns. But the kicker is that about once every 4 years he'd lose 75% of his account value, and there is even a small possibility of a complete 100% loss. Even with those outlandish returns, it isn't worth it for most people, so they are wise to stick to max drawdowns of less than 25% like you suggested. In my experience, most people who really understand risk prefer to limit it to less than 15%.

I'm excited that you've posted a real example as I've been following along but pretty confused about how one might calculate expectancy.

I have two questions:

1. You say that to triple the risk, I triple the trading size. But the risk is all expressed as a percentage. Do you mean that I lever by a factor of three? Surely the percentage loss on the trade isn't represented as a percentage of total portfolio? Or is your first example assuming a standard percentage of the portfolio is put at risk, so the actual percentage losses are percentage of the portfolio?
2. How in the world would I generate these beautiful numbers that I would then do the math with? Best I can do is look at a bull flag coming into the 61.8% fib, place a stop below the fib a ways, and then give myself a ratio of distance to loss vs the expected price target (AB=CD). Even that's really a lot of guess work and price targets are a bit of voodoo.
   
   As an engineer, it feels a little bit like you may be confusing accuracy and precision. You have these very precise numbers with lots of decimal places, but aren't they kind of shots in the dark? Or are you taking these from a long term trend/system looking at reams of historical data and giving you the likelihoods based on all that data? If so, then it's probably outside the reach of your average retail trader.

By the way, thank you for a very enlightening discussion. I think there are quite a few cruisers who use or hope to use some sort of trading as at least a supplement to their cruising budgets or as a means to build the kitty in the first place while we're land-bound.

[Certeza]

Originally Posted by cwyckham

I have two questions:

You say that to triple the risk, I triple the trading size. But the risk is all expressed as a percentage. Do you mean that I lever by a factor of three? Surely the percentage loss on the trade isn't represented as a percentage of total portfolio? Or is your first example assuming a standard percentage of the portfolio is put at risk, so the actual percentage losses are percentage of the portfolio?

Yes, I meant triple the leverage, which is the same as thing as buying or selling three times as many shares or contracts. Yes, the % loss is being expressed as a % of the total portfolio. Calculating everything based on the effect on the entire portfolio cuts out unnecessary steps.



For example, let's say in your $50K account you traded a futures spread and the broker required $1,000 margin. Assume you get a gain of $1,000 on that trade. Was your return 100%? Well, to make yourself feel good you could boast about your 100% return on margin all day long, but at the end of the day you still made $1,000 on a $50K account, which is only a 2% return. All available capital for trading, even if not presently held in the account, must be considered for determining true performance.



Do yourself a favor and think of every trade based on how it affects the entire portfolio. And get in a habit of thinking in percentages. That way, as you scale up your account you don't have to adjust the numbers bigger as the account grows.

How in the world would I generate these beautiful numbers that I would then do the math with? Best I can do is look at a bull flag coming into the 61.8% fib, place a stop below the fib a ways, and then give myself a ratio of distance to loss vs the expected price target (AB=CD). Even that's really a lot of guess work and price targets are a bit of voodoo.

That feeling comes from the types of indicators you are using. Technical indicators have statistically poor reliability. Thousands of people are constantly trying to come up with the best combination of these unreliable indicators that might result in the realization of positive expectancy. My trading has nothing to do with bull flags, Fibonacci, candlesticks, etc. My opinion is that the only reason those indicators ever work is because they are sometimes self-fulfilling. IOW, if a whole bunch of people believe that the 61.8% fib will hold, then they will all pile up their buy orders there and it will hold simply because they all believed it would. VooDoo is a very apt description in my opinion.

As an engineer, it feels a little bit like you may be confusing accuracy and precision. You have these very precise numbers with lots of decimal places, but aren't they kind of shots in the dark?

No it isn't confusing accuracy and precision, and no they aren't really shots in the dark. There are only two types of participation in the markets. In the first you buy into a company and wait for either the strength of the company, strength of the economy, or inflation to result in a higher price. That strategy should always be long term buy and hold. EVERYTHING else is statistics based. If a trader doesn't have a firm grasp on statistics, he should go back to buy and hold investing until he does.



Understanding that everything else is statistics based, I can explain further. There was once a saying originating from a old TV show, "it's bigger than a bread box, smaller than a ______". The idea is that when accuracy is impossible you begin with the extreme on either end and work your way in to becoming increasingly accurate. XXX is currently trading at $20, what will the price of XXX be in 20 days? Start with extremes that are a virtual certainty, like the price of XXX won't fall below $0 or go above $1,000, and work your way in. Using all available statistical data, you can determine the probability of any event happening in the future, if XXX behaved similar to the past. Then obviously you must add in the probability of any unprecedented event. You end up with a statistical distribution that has tails that reach out further than the historical distribution to account for the unknown.



Admittedly, doing this for stocks tends to result in very wide distributions that are often useless. This is because stocks are subject to event risk. For example, regardless of what a stock has done in the past, there is no assurance that tomorrow we won't uncover a massive fraud that halts trading on the stock and results in a $0 price. That's one reason why I trade futures in a manner that I can dramatically reduce, and in some cases completely eliminate, the statistical likelihood of an event shock. By eliminating the extreme negative events, you start to get a distribution that is useable.


In the end, when done properly you have an accurate representation of the probability of any future occurance. That's why in my example I couldn't say that a loss greater than 4% was impossible. I just said that the probability of it happening was 0.0001%. When you start out, your models are wide due to greater margin for error, and as they become more robust they get more narrow and more useable. From there, expectancy takes over and you construct trades based on those concepts.

Or are you taking these from a long term trend/system looking at reams of historical data and giving you the likelihoods based on all that data? If so, then it's probably outside the reach of your average retail trader.

In my particular case, I have models that consider 70 years worth of historical data. So yes, even with a powerful computer it takes a while to run a full analysis. But keep in mind that when I developed my models, I did so as a retail trader. Was I the "average retail trader"? Not sure. I did have knowledge of and an affinity for higher level math and stats.



To this day, the models are getting better all the time. For example, in 2013 there was a condition in the markets that I work in that had never happened before. The existing models couldn't handle it well because there was no historical data to draw on. I could've just plowed ahead blindly just chasing returns, but the prudent thing was to contact clients, explain that we'd be scaling back for a bit in order to avoid exceeding risk limits, and then working feverishly to add to the models so they would work in the new market conditions. It ended up taking longer than I'd hoped, but every good trader should be willing to stay within the set risk parameters and never take uncalculated risks.


The thing about what is or isn't outside the reach of a retail trader is that nothing is technically outside your reach, but you can't expect that if you are starting from zero right now, next month you'll have something that took me 15 years to create. I my experience, the best programs aren't for sale to the retail trader. They are developed and operated by professionals and are pretty closely guarded. There are literally thousands of mediocre systems for sale for either a small subscription or purchase.



That is the crux of the argument on whether to hire a professional. If you are determined to become a great trader, then work hard at it. Try not to re-invent the wheel. If you have easy (CHEAP) access to a system that works, purchase it to take advantage of the work the developer already did, then dissect it and try to perfect it. But if your goal is simply to get the best returns for your risk level, you might consider the route of hiring a professional. I mean, let's say you have a 25% risk tolerance and the best system you can find averages 25% annual returns for that amount of risk. Maybe you feel that you could improve the system to 35% returns within 5 years on the same risk. Now assume you had access to a professional system that for the same risk averaged 40% annual after fees, and you didn't have to do any of the work. Seems like a no-brainer to me.

{edit} just so there isn't any confusion, and people don't jump down my throat. I AM NOT implying that the suggested professional would be me, or that I fit that description.

[cwyckham]

Certeza,

I've very much enjoyed your posts. They are very well thought out and have introduced concepts I haven't really thought about before. I'll be honest, I think in terms of stocks, so maybe that colours my thinking a bit, but I don't understand how 70 years of history necessarily tells me what this particular stock is likely to do in the next month.

One area that I think that you and Delfin disagree on is what a drawdown means. Delfin is comfortable with a 25% drawdown because he's confident that it is rare enough and will be more than offset by future gains.

Likewise, even an index etf like SPY has high volatility compared to other types of investments, but it also returns much more in the long haul. You almost need to add a time dimension into the risk analysis. I think you're saying your clients won't stand for a 5% drawdown at any time and Delfin is saying that he will accept an instantaneous drawdown that is closer to 35-40%, but over a 2-3 year period, he insists on a net upward trajectory.

[Therapy]

Originally Posted by Delfin

I think one of the best ways to learn is to setup a play money account at Interactive Brokers, take a free trial for 30 days of Moore's service and play with it. You'll learn what you need pretty quickly, although I have to tell you, figuring out Interactive Brokers is more of a headache than getting your head around spread trading. There are many investing alternatives of course, one of which is to turn your money over to someone with a vested interest in convincing you that you're too stupid to do this yourself. Such is the world of investment advisers. Best of luck.

All I saw for "play" account was for a teacher for their class. Had to prove they are a teacher.
Did I miss a button?

[Therapy]

Originally Posted by Certeza

[/INDENT]To extend it a little further, if he was willing to risk 75% drawdowns on trades like this, he'd be looking for at least 357% annual returns. But the kicker is that about once every 4 years he'd lose 75% of his account value, and there is even a small possibility of a complete 100% loss. Even with those outlandish returns, it isn't worth it for most people, so they are wise to stick to max drawdowns of less than 25% like you suggested. In my experience, most people who really understand risk prefer to limit it to less than 15%.

If someone was willing to risk 100% of 50k in one year for 357% returns, would that be unwise.

[hellosailor]

The key thing to remember about any stock market, or commodities exchange, or similar, is that it is a CASINO, not an investment or financial institution.

Logic does not drive them, emotion does. And you'll hear business news discussing the "fear index" just as much as any other factor. When people get frightened and upset, say, by discord in the Ukraine, or Congress being brats...the market plunges! Why?? Simply and solely because people are upset and they ignore the real questions, like whether a company will be making profits and growing, and they start trading and bailing out of things instead.

Lots of people have a system, sometimes it works for them and works for a long time. But if "systems" of any sort really worked 100% reliably, or even 51% reliably, over the long term, then there'd be a lot of billionaires managing those sure-fire-win funds.

And there aren't. The folks making money, are taking their percent--a fixed flat small percent--on the money they manage for other people. They folks making money, prefer a guaranteed percent of each trade, or a guaranteed percent of each portfolio, because they know that's the only long-term guarantee.

Which is not to stop anyone from investing in a winning system, just to say that you've also got to know when to bail when the system isn't winning any more. And most of all, remember it is a CASINO. The house (the brokers) always win. Only take in what you can afford to lose, because something like 95% of the market loses all the time, in order to give the other 5% a profit.

If you're careful and not greedy...sometimes you can win.

[Certeza]

Originally Posted by Therapy

If someone was willing to risk 100% of 50k in one year for 357% returns, would that be unwise.

Yes, for various reasons. The main one being that there is no telling when that 100% loss is going to happen.

If that was a person's entire savings, it would be insane to take a risk like that. It is completely non-scaleable, because as mentioned in the example, once every 4 years you'd blow up.

If it is only a small fraction of your total portfolio then you aren't really getting a 357% return. Allow me to illustrate...

Alternatively, let's say you have $1MM in investment accounts, and you say to yourself, instead of taking limited risk with the entire portfolio, I'll just pull out $40K and take 100% risk with that. Then if I lose 100% I'll just pull out another $40K and try again. What this really means is the following;

$1,000,000 accessible portfolio
$40,000 risk
$142,800 return

In other words, 4% risk to get 14.28% returns on your portfolio. That's what I've been trying to explain. Any capital held outside the account that is accessible if needed is called "notional capital" and is included when determining performance.

Now there are absolutely ways to use notional capital to your benefit, but that is beyond this discussion.

[Certeza]

Originally Posted by cwyckham

Certeza,

I've very much enjoyed your posts. They are very well thought out and have introduced concepts I haven't really thought about before. I'll be honest, I think in terms of stocks, so maybe that colours my thinking a bit, but I don't understand how 70 years of history necessarily tells me what this particular stock is likely to do in the next month.

One area that I think that you and Delfin disagree on is what a drawdown means. Delfin is comfortable with a 25% drawdown because he's confident that it is rare enough and will be more than offset by future gains.

Likewise, even an index etf like SPY has high volatility compared to other types of investments, but it also returns much more in the long haul. You almost need to add a time dimension into the risk analysis. I think you're saying your clients won't stand for a 5% drawdown at any time and Delfin is saying that he will accept an instantaneous drawdown that is closer to 35-40%, but over a 2-3 year period, he insists on a net upward trajectory.

Well that is going to be one of your problems. You are correct, 70 years of history has nothing to do with an individual stock. Individual stocks contain individual risks and individual price action. My advice, don't trade individual stocks. Problem solved.

No really, all joking aside, I've run intense analysis on stock trading and stock traders, it is VERY rare that a retail trader in that arena is adequately compensated for risk.

re: the definition of a drawdown...

You are unwittingly referring to the difference between Graham's risk and Markowitz's risk. The former basically says that anything you'll likely recover from isn't actual risk. The latter basically says that recovery cannot be guaranteed, and therefore risk is defined by the volatility of your drawdowns.

I'm not gonna argue who is correct, except to say that there really is not (and cannot be) a guarantee that any program will recover from any given loss. And that becomes increasingly true the larger the drawdowns get.

For example, a person can easily argue that a system with a max drawdown over 20 years of 5% is very likely to recover. The same cannot be said for a system with a max 80% drawdown, because after a 80% drawdown you require a 400% gain to get back to even.

When it comes down to it, none of that really matters to the preceding discussion. The risk analytics and expectancy figures already accounted for the likelihood that a system would recover because the gains from the recoveries were already factored in. At this point, the numbers on that system all basically say that in comparison to the risk, the returns would be considered 'decent'. My opinion is that particular system would be best suited for an account size of at least $120,000, still trading 1x1 spreads. In that case you'd be risking occasional 25% drawdowns to realize an average 29% return.

But I should note that according to the posted equity curve for that system, on a $40K account the drawdowns would be FAR greater than the 25% you suggested and not anywhere close to being rare. On the contrary, that equity curve indicates that on a $40K acct, 20-30% drawdowns are to be expected every year.

[Therapy]

Originally Posted by cwyckham

By the way, thank you for a very enlightening discussion. I think there are quite a few cruisers who use or hope to use some sort of trading as at least a supplement to their cruising budgets or as a means to build the kitty in the first place while we're land-bound.

+1 on that. I think if anyone has any money saved anywhere they want to see some sort of return. Obviously, we're talking a lot about risk year, but I know that all of us want to have some sort of return and unlike the old days, The money markets and CDs of today are woefully inadequate.

[frank_f]

Originally Posted by Therapy

The money markets and CDs of today are woefully inadequate.

+1 there.

My company matches my pension contributions and invests it with a money market manager. After almost 15 years I have about a 1 increase over the contributions. The poor returns didn't stop them from taking their fees every year and it doesn't seem like I can move the money to another firm. I could have done better with a savings account.

[Therapy]

Originally Posted by Certeza

Yes, for various reasons. The main one being that there is no telling when that 100% loss is going to happen.

If that was a person's entire savings, it would be insane to take a risk like that. It is completely non-scaleable, because as mentioned in the example, once every 4 years you'd blow up.

If it is only a small fraction of your total portfolio then you aren't really getting a 357% return. Allow me to illustrate...

Alternatively, let's say you have $1MM in investment accounts, and you say to yourself, instead of taking limited risk with the entire portfolio, I'll just pull out $40K and take 100% risk with that. Then if I lose 100% I'll just pull out another $40K and try again. What this really means is the following;

$1,000,000 accessible portfolio
$40,000 risk
$142,800 return

In other words, 4% risk to get 14.28% returns on your portfolio. That's what I've been trying to explain. Any capital held outside the account that is accessible if needed is called "notional capital" and is included when determining performance.

Now there are absolutely ways to use notional capital to your benefit, but that is beyond this discussion.

Okay. Stated that way, a 4% risk for a 14% gain. Sounds pretty good to me.

[Certeza]

Originally Posted by frank_f

+1 there.



My company matches my pension contributions and invests it with a money market manager. After almost 15 years I have about a 1 increase over the contributions. The poor returns didn't stop them from taking their fees every year and it doesn't seem like I can move the money to another firm. I could have done better with a savings account.

Yeah, not a big fan of that type of manager. IMO, charging a 1/2% management fee to do nothing and only get your clients 1% annual is robbery.

If a professional is gonna charge, he should be offering target returns that are above what the client could do themselves on the same level of risk.