Introducing Thor

Thor is a daily chart pattern trend following system which Scott came up with based on the work we both did for Heisenberg, and which is further leveraged off the old work done by Ivan. It is, in essence, an ‘Ivan take’ on a classic trend following system of the type everyone knows and was spearheaded by Richard Dennis and William Eckhardt (i.e the founders of the turtles). However the campaign management has been tweaked to produce a high SQN trend trading system, currently we are measuring approximately an SQN of 3.5.

High SQN equates with low drawdowns, as opposed to classic Donchian breakout trend following which has massive winners but massive drawdowns. It only trades during strongly trending price action on multiple timeframes, and Scott and I have exhaustively tested custom trend filters until we recently came up with what we believe is a genuine innovation – and which has resulted in dramatically improved performance this month.

We will not be discussing these filters, it is proprietary and frankly we don’t want anyone to figure it out. The campaign management you could obviously reverse engineer, but the rest of it will be very difficult but not impossible to reverse engineer. There is also a significant body of work in selecting which pairs and which futures contracts to run this over at any one time, and we have a fantastic model for this, again secret sauce proprietary.

The bottom line for those currently racing to reverse engineer our system (we can’t blame you for it) is that you will gain significant value from just paying for it. If we hadn’t build it ourselves we would pay for it, and that’s the honest truth. It requires you to be at your screen from 5:00 pm-5:15 pm EDT for currencies and 5:45pm – 6:00 pm EDT for the CME futures (metals, energy and index futures), the grains open 1 hour after that.

The Basics

Initial capital requirements are minimal for FX trading (you could trade with as little as $5k) and to add all of the futures contracts in I would say a minimum of $25k, but even $25k would be a bit restrictive depending on your risk tolerance. Certain contracts are quite large and a typical R value on ES futures would be roughly $500 per contract (could be higher or lower) and a typical R value on CL would be $1500, GC would be $700-1100, SI might be $2000.

Scott currently trades a small account at a high R value for income purposes (withdrawing the money each month to live off) and I am getting ready to do the same. Scott compounded it quite easily from $28,000 to $50,500 at close Friday in 6 weeks trading at $1000 R values (quite unacceptably high at first). He also trades a larger account at a 1.5% R value with a negligible risk of ruin which he has compounded from $600,000 to $860,000 (43% return) in just under two months. Yes you read that right.

The system is performing to specifications currently, not above or below, just exactly as it should. The backtested win rate is 57% and the current win rate is 56% on closed trades (though that will go up slightly when we bank winners from some of the open trades) My guess is that this is an exceptional period we are having due to the USD breakout, but something that should happen enough that it is not surprising. We would be extremely happy with long term performance 75% of what we are currently doing, and think that is a reasonable expectation.

Deviating from the rules, not being at your computer when you are supposed to, holding on to winners/losers when we tell you to jump out, not having stops set correctly, taking a different correlated pair than we tell you (for example let’s say we have a setup in gold and silver. We tell you which one to take but you like the other one) is going to hurt you over time. The system requires a serious commitment to trading. It is for people who want to compound an account significantly and quickly, shooting for 400%+ returns, and are prepared to work hard to do it. Scott’s personal goal is 400% per annum, and he thinks that he has maybe a 50% shot at it. Then again, he’s a bit crazy and he’s Australian – so don’t hold that against him.

Your Commitment

If you cannot free yourself to be at the screen during those times I would advise you to choose a different system. You do NOT have to be at the screen at other times, and to a certain extent screen watching is counter productive. Bottom line is that if you have a small account you will be limited to what you can trade, which is fine. There is plenty of opportunity. The system takes approximately 30 trades per month the way Scott trades it right now (manually), which is to avoid correlated pairs and have strict risk requirements about the number of simultaneous trades which are not yet at break/even.

You may be less risk averse, or take more trades. For instance if you get a long signal in NQ, which is in profit but your stop is not yet at break/even, and you got another long signal in ES the next day, you may choose to take the second signal as well. Personally we don’t do that, or anything approaching that. The Achilles heel of every daily chart system is that correlations are in reality orders of magnitude larger than mathematical models suggest, so you must trade accordingly. The bonds are related in some weird way to the sugar futures which again are somehow related to GBPJPY. We don’t know why and we don’t care – we just don’t correlate them. Also the setups cluster – you may not get anything for 2 days then 5 in one day, and any setup you instigate on the same day is also correlated in a way difficult to quantify. So if you don’t get a signal for one day please do yourself a favor by not emailing me. For my wrath shall wax hot, your wives shall be widows, and your children find themselves fatherless.

The Good

What sort of return can you expect? In the 7 years of backtesting the WORST periods the system did expectancies in the 0.25 range. The best periods were whole years for certain symbols in the 0.8 range. Long term backtest average is 0.5 expectancy. The bigger the trend, the higher the return – and yes, it works very well during bear markets. Thor would have killed it in 2007/2008. This system does not work in any way shape or form in sideways markets, but our filters keep us on the sidelines as much as possible. We are trend hunters, seeing the strongest trending markets worldwide to trade.

If you take 30 trades per month at .5 expectancy and 2% R value, on average you will make 30% per month if you trade perfectly. Yes you read that correctly. If you take 30 trades per month at 2% R value and 0.25 expectancy you will earn 15% per month on average, less mistakes from you deciding you know better than the system. That’s on you. At 30% per month you double your account in just under 3 months. At 15% per month you double your account every 5 months, or 535% per year. Frigging crazy numbers, but it’s not magic. Takes a lot of trades x significant edge = hell a lot of cash.

The Bad

How hard is it to trade correctly for a whole month? Very very hard. The email alerts will make it easier. Scott personally made two mistakes this week, the first for him in 50 days. Your head will screw with you in a hundred different ways and try and get you to not follow the system. We are here to help you follow the system. To help him follow the system Scott actually employs someone he trusts to keep him on track. This way he doesn’t just decide to go off the reservation on his own. It’s a constant dangling sword of Damocles and we all are only a day away from turning back into a nervous little mess like anyone else. I would budget for 2-3R of mistakes per month unless you are already an expert, and obviously that can seriously eat into your bottom line.

It’s very easy to trade one single day perfectly. All of you could do it, not a problem. What is difficult is doing it day after day after day. Also, it’s quite easy doing it on a small account, but as the numbers get better psychological issues can come into play. That is why we will always think and talk about R and not dollar values. Except when it comes to the profit graphs – Scott will continue to post his results in $$ values as they are up to date. Thor does stay in campaigns for weeks on end sometimes and the Google spreadsheet will not always be in sync as it only logs completed campaigns.

And The Ugly 

What sort of drawdowns can you expect? The largest drawdowns we see in backtesting are 6R. A rule of thumb for me is to double what you see in a significant backtest, so you can reasonably expect a 12R drawdowns at least once a year, and budget on getting it twice. In my opinion this system contains a significant black swan risk from correlations, so once or twice a year you will have 10 positions open and get butt-slammed on most of them, so your 12R drawdown may very well happen in the space of a week.

What sort of R value is appropriate? That depends on your risk profile. If you don’t wish to experience more than 5% drawdown you must therefore trade with something like .5% R value which will give a nice smooth equity curve at around 7% per month less mistakes. You might decide that you have a different risk tolerance with your own money and the market’s money. In that case you might trade small (.5%) until you have a 30% gain, and then be willing to trade an R value that you would risk all of that 30% at a higher R value of 2%.

A Fistful Of Dollars

Whatever you decide to do it is critical you have a business plan for what you do and when you do it all typed up and laminated and stuck to the side of your computer. Trust us, when you get close to your uncle point you won’t want to stop, and you will be at your least objective and incapable of making a decision. You MUST decide at what point you stop trading.

Scott’s personal preference in trading plans is to decide to reduce risk the further you get in a drawdown, so that you approach your comfort limits (where by definition your trading will be at it’s worst since trading in a drawdown is way harder than winning all the time) asymptotically. An example plan might be:

1. Trade with 1% R value until equity is up 20% AND you are trading at 95% efficiency for the last month (less than 1 mistake in 20 trades)
2. When efficiency is > 95% and SQN > 3.0 and equity 20% above starting capital trade 2% R value.
3. At 12% drawdown from peak equity, measured in $ value from peak, halve size to 1% R value
4. At 15% drawdown from peak equity measured in $ value from peak, have size again to .5% R value
5. At 20% drawdown from peak equity measured in $ value from peak, take a week off from new positions in the market, managing only existing positions and evaluate whether your edge exists of whether you are trading at acceptable efficiency.
6. If you recommence trading after a week break at 20% drawdown lower size to .25% until you can trade 20 trades without making a mistake
7. Any time you are below your original capital trade trade for .5% R value until you are 10% below original capital. At 10% drawdown from original capital lower position size to .25%R value. At 15% lower position size to .125% R value.

You could also add what Scott calls the ‘two bucket’ plan above a certain sum of money you never thought the market would give you. Let’s say that you decided on the business plan above. You could also put in that any amount > 50% you earn in a year goes into a separate bucket, the ‘roll the f..ing dice bucket’. In this bucket you trade everything at 5% R value (which is going to have you periodically maxing out available leverage and undergoing significant and real risk of ruin for that bucket. The above is to illustrate that the actual system (though good) is the smallest part of the plan. You achieve your financial objectives with position sizing and you safeguard yourself with a business plan.

The Ground Rules

Over the past few weeks Scott and I have actually been debating about whether or not I should offer Thor to any subscribers. For one we believe that only a small minority of you will be able to follow the system properly, plus several of the contracts we trade require decent size accounts (again, Forex is fine even for small cracker accounts).

Someone actually asked me in the comment section why we don’t keep it for ourselves if it’s really so great. And he’s right about that but that argument obviously would apply to the entire blog as a basis to questioning my profitability or credibility. However, after six plus years and over 3000 posts I think my track records speaks for itself. And psychologically it’s beneficial for me not having to rely on trading income completely as it gives me a better sense of detachment. Plus it’s been a lot of fun for me and Evil Speculator’s premise from the get-go has been one of helping retail rats avoid the trials and tribulations of trading the markets and of course bank some serious coin. If I stop doing that or if you aren’t successful then you’ll obviously walk away and in the end so will Evil Speculator.

Limitations

However, that said, there will be limitations to how many people can sign up to Thor and in particular regarding the jumping in and out of it. We’ll do our best to help you guys bank coin with Thor and keep you in sync with how we are trading. For instance if we take discretionary entries that bend the core rules a bit or if we skip one setup for some reason (FOMC, some market event, etc.) we will let you all know beforehand. As you can imagine that is a commitment on our end and we are only prepared to do that for serious and equally committed participants. What I do not want to find myself doing is a lot of hand holding and explanations for people who jump out of the system two months later because it may hit 10R draw down (or especially if they screw up on their own). It’s quite an investment keeping you guys on track, frankly speaking Scott has repeatedly advised me to NOT do it. I still see value but if it becomes to much for me to maintain I may disable signups. To preempt any concerns – anyone grandfathered in will be retained as long as the signal is running – even if we decide to disable new subscribers.

This Is Not Sparta

There will never be more than 100 subs for Thor – I think that is the limit both Scott and I feel comfortable with. I will also refuse anyone who’s quite obviously working for a bank, fund, or prop desk. I hereby reserve the right to reimburse any Thor subscriber and cut that person off if I see the need. And trust me – if we see large cars moving along or with our entry signals (or attempts to fade them) then we will find the responsible party and ban them for life. We have the skills to do that and will employ a zero tolerance policy.

The Cost

I have given the sub fee a lot of thought and decided to offer a special intro rate of $149 for the first three months – if it keeps plugging along as it has recently I’ll raise it to $199 in early spring. The folks who sign up now will get grandfathered into the old price for life – unless their subscription expires due to non-payment and PayPal kicks them off.  IMPORTANT: Anyone who signs up and cancels will not be able to sign up again – it’s a one-time opportunity. I can control that on the aMember side very nicely. W have zero patience for retail guys who go on vacation for three weeks and decide they don’t want to pay their subscription during that time. You may laugh now but I even get Gold subscribers who cancel their recurring subs every time they take a few weeks off – still enjoying the recurring price but treating it like a non-recurring one. Yes, we get all the high rollers here at Evil Speculator.

CrazyIvan Subs Rejoice

And finally, I’ll be giving all CrazyIvan subs a three month freebie – they’ll deserve it as CI has been pushing sideways for six months now and we feel their pain. After that they’ll get the chance to sub in for $149 before I’ll raise it to $199.

I think all of the above is pretty fair – if not, well, just don’t sign up – it’s all the same to us. We’re not implementing these limitations because we’ve got a chip on our shoulder but because we need to limit Thor’s exposure. But we all know that a high SQN system like this is going to make a bit of a splash, in particular if we keep posting a rising equity curve week after week, month after month. Which is why we will limit this to our core rats and early adopters. Obviously if 1000 people and a dozen fund managers start trading Thor then performance will suffer and we all lose in the end.

If I left anything out please post it in the comment section. You can sign up for for the free beta right now. Depending on the issues I see it’ll last between one to two weeks. I probably won’t deploy Thor until tomorrow or Thursday but I recommend you do it now as I will pull the free beta sub in a few days.