Stock, Algorithms

Why our Stock Algorithms are Different Than Most

“Tell me about your algorithms for trading stocks and how you are achieving these winning results?” September 23,

“Tell me about your algorithms for trading stocks and how you are achieving these winning results?”

September 23, 2019 Update: White Paper: How EPIC v3 Crude Oil Machine Trading Outperforms Conventional Trading Methods

Feb 20, 2019 algorithm development article here:

That was a recent question posed by a reporter during a telephone call in preparation for an upcoming show I am booked to appear on. The reporter’s question caused me serious pause… the “how and w hy” started to consume my thought process. And, there have been many, from friends and family to our subscribers that have asked similar questions. I get it; what causes someone to dive deep enough to actually do that?

The answer is really quite simple, “Necessity is the mother of invention.”

So here is a bit about my journey so far, how the algorithms were developed, and how they may or may not help you with your trading.

I Needed to Win – The Market Changed.

I trade. I have been trading since I was twenty-one years old (so almost thirty years). But recently, the stock market has changed in three distinct ways that make it more difficult for a trader to always operate in the ultimate position, being “the trader’s edge”.

  1. Computer algorithms (Jim Cramer commented on this topic recently on Mad Money on CNBC).
  2. The Federal Reserve monetary policy, this point being the most important as far as I’m concerned.
  3. The sideways markets the last few years.

So I set out to solve the problem for myself: how do I get back my trader’s edge? And if I can’t, then I shouldn’t trade.

Most Stock Market Algorithms Do Not Work or Are Too Expensive.

Yes, it is true. Similar to how 90% or more of day trader’s fail, at least 90% of stock market algorithms either do not work or are outright scams preying on people that are failing. Maybe they had good intentions when they set out to develop their algorithms, but most do not work.

The majority of these algorithms are high frequency or FX market-type automated tools (“bots”) that represent some form of winning percentages to the public. However, the truth of it is that the successful algorithms, the real good ones that actually do perform well, are simply out of reach to the retail public. A bottom of the barrel algorithm costs minimum 4,000.00 per month and I have looked into some (available at a retail level) upward 200,000.00 per month or more. If any of these “out of reach” algorithms are even made available at all, you will pay dearly. And why not, I suppose? If they work they’re worth it, especially algorithms that can hit 80%, 90% or more.

The problem for me was, I couldn’t afford upwards 200,000.00 per month for a decent algorithm. I am a doubter by nature, and price tags within these ranges put my hard-earned money at risk. So, I set out to figure the mechanics out on my own.

My Mission: Develop the Math and Start Making Public Calls.

While I was in the Caribbean with my family last winter, I started to work on the math. When I felt I was getting close, I started to publish calls on Twitter for the six algorithms I was working on (in the quietest way possible, but still making calls so it was public and undeniable if the math actually worked – which I doubted by the way, so I was really extending myself there). The six algos I started to work on are Oil FX: $USOIL $WTI, $SPY (S&P 500), $GLD (Gold), $SLV (Silver), $DXY (US Dollar Index), and the $VIX (volatility index). I have also worked on natural gas – but the math is “off” so I don’t know that I will ever publish it.

Over time, the calls I was making (based on the algorithms and not my personal trader bias) started to hit. I was working on different time frames from intra-day, to swing, and months out… and they’ve all been hitting at better than 90% (the tighter the time cycle the higher the probability of a win hit). When my own trader/human bias was involved, I was lucky to still be hitting 60%. This was the most difficult and humbling part of it for me, the realization that my mind as a trader couldn’t outperform the math.

“How can simple math beat me, the ultra, omnipotent trader?”

Seven months later, and I still struggle with trusting the algo calls (the simple math), but as time moves on, I’m having less difficulty with this.

So What Are These Algorithms?

Simply put, these algorithms are based on:

  1. Traditional math (simple logic), which involves considerable weight toward simple average.
  2. Traditional algorithm modeling disciplines.
  3. Traditional stock market charting and indicators.

It is a mixture of these three components that constitute what the algorithms we use are based on.

The most important thing to understand is these algorithms are not high frequency/bot or “automated-type” algorithms. And they are not cryptic balls using a “crystal ball”; they are scientific and represent simple math. There is no crystal ball, no “top-secret” artificial intelligence, and no geopolitical reasoning.

Visit this link for a list of the most common algorithms found/used on the stock market.

My algorithms are different, in that they are probability algorithms based on the most absolute logic available, designed with the goal in mind to provide the trader (whether it be for intra-day, swings, or investing) an edge in a specific stock, currency or commodity by representing the conclusions of the math on a traditional 2D stock chart. In other words,

Keep it Simple Stupid! If you can’t represent the conclusions of the algorithm on a chart that a trader can use to trigger their own trades, then it is useless as far as I am concerned.

The more these algorithms can be used in traditional charting and on similar platforms, the better. They are, after all, developed by taking traditional charting indicators (that are represented on a chart) to start with. So why not? I suppose there is an argument for the high frequency bots and other tools in my category – but for my purpose, that wasn’t my goal or intended use and purpose.

So think of it like this: when you learn how to use the Fibonacci retracement indicator, for example, it is represented on a chart in a specific way. That is, in its simplest form, what we are doing. We are representing our algorithmic indicators in a specific way on a chart for the trader to use. Fibonacci is an algorithm in itself, as are the other indicators traders use everyday (most traders don’t think of them that way – but they are). Its just that we are taking what we discover to be the best indicators for specific instruments and extracting the best probabilities from a group of indicators and representing that as a probability to the trader – on a chart, in a specific way. It turns out, based on conversations with software developers I’ve had, that this is a very technical process called “reduction”: representing one problem as something “just as difficult” or “easier” than a complex one.

Here is an interesting Ted Talk video that puts a perspective on it that is more similar to what our work objectives and methodology involves:

More Specifically, How Is Each Algorithm Processed?

The easy part is the math, but representing the mathematical conclusions on a 2D chart in such a way that is easily usable by a trader is the hard part. The math is simple (in that the math and charting is standard and nothing crystal ball like – it is, a scientific process) but once you have the results it then becomes how the heck you get that on a chart – and for all the different time frames. The real challenge is how to project the results of these mathematical processes onto a chart, for all the different time frames, that is also human-readable and “intuitive.”

In its most basic form, the development process can be detailed as follows:

  1. Simple Averages: Stock charts present endless opportunities to run averages. What is the average price annually? Last 4 years? Last month? What is the average price of crude at 10:30 Wednesday morning? What is the average drop in crude at EIA report time? Average spike? What is the average spike in the S&P at 2:30 PM? And on and on and on and on. So for each algorithm there are hundreds of averages or patterns as a result of averages – or better described as probabilities.
  2. Indicators: Take the traditional charting for the equity, currency or commodity you are working with and use each time frame you want to work with and determine which traditional indicators work better than others. Then weigh each indicator in accordance to its “win rate”. Then take that data and relate it to step 1 above. Simple right? Simple logic if you ask me. Different equities, currencies and commodities trade different in relation to various traditional indicators. So it’s just charting on steroids. But – using simple logic.
  3. Modeling: This gets a little more complex. Then you take the various traditional algorithmic disciplines that you understand can be applied to stocks and begin to run simple models (patterns, averages, timing, price in relation to time, etc). One very important part in this is removing the anomalies. In other words, every stock, currency or commodity will have anomalies, which take it out of (remove it from) its natural trading pattern. For example, Fed talk affects the S&P 500, or currency is affected by currency wars and oil is affected by rig counts and inventories. So you remove the anomalies and you work with your modelling. This is where you get your quadrants from by the way.

Then, you need to take this data and represent it on a traditional chart so that a trader can use the information to gain an advantage in specific scenarios (in different time frames for swing, daytrading and investing) – the trader’s edge.

This is the hard part – representing all that data in an easy to understand way for the trader on a chart.

So How Do I Use The Algorithm?

Each algorithm is charted as I mentioned, so it becomes a process of understanding how to use the algorithmic chart indicators (that we develop and provide) to your advantage. A quick visit to an EPIC the Oil Algo blog post may help understanding (keep in mind one post won’t show all the indicators because they are a running story – but you will get the point). It really comes down to time frames of trade for swings, investing or day trading and the specific indicators. The primary indicators we provide our subscribers are (and they are growing):

  1. Time/Price Cycle: This indicator is proving to be very helpful. You will find on my personal feed, or for example on Epic’s Twitter feed, how absolutely and incredibly accurate these have been. Time/price cycles are important because they signal a change in trend – and knowing when a trend is going to change is the best edge a trader can have. This has been my number one edge because I scale into trades. I day trade when a currency, commodity or index stock is at an inflection point with the objective being to get on the right side of the trade. Once I am on the right side of the trend it is hammer down time for me – and truth be told it is vital because you only get so many chances in that over a five year period for each commodity, index or currency. We provide these as written times of the week (day and hour) for our traders.
  2. Alpha Algo Targets: Targets are great. Having targets that hit with regularity are even better. Our algos are hitting targets at 80-95% depending on the algorithm and time frame you are looking at. Calls days out we are hitting between 80 – 90%, calls months out we are hitting in that range too (but our algos are only seven months old so data is difficult to brag about) and intra-day we are hitting well over 90% with most of the algos. If you review an EPIC member blog post you will see these as red circles on the chart.
  3. Alpha Algo Trend-Lines: These are trend lines just like traditional trend-lines that are established primarily as a result of averages (taking into consideration time/price cycles) and how the price of the equity interacts with price. In other words, algorithms out there are using averages to such a degree that we can actually determine where the lines are because price is affected when the price is traded across the line upward or down. These anomalies in price action are a result of machine trading. Why is this important to know? Because the trend-lines act like traditional trend-lines in that they represent support and resistance. If you review an EPIC member blog post you will see these as red dotted lines on the chart.
  4. Algorithmic Trading Quadrants (or trading ranges): Quadrants are more complicated to explain in short, but like described in the Ted Talk video above, we have discovered quadrants or geometric shapes in which stocks will trade, most specifically within large liquidity environments such as with currencies, commodities or indices. The quadrants are represented on our charting for our traders only when they are predictable and provide an edge. When they are in play they are precise to say the least and provide a trader with pin-point sniper intra-day trading (because you are in essence trading along with the machines). These quadrants can be intra-day or even represented on up to 5 year charts we have discovered. Great examples of these wider time frame calls would be our calls with the US Dollar, Silver and Gold – nobody believed the calls we started making months ago and all of our price targets have recently hit – it blew people away.

Here is an example algorithm represented on a 2D chart:



Target Called Days in Advance! On Fire! $USOIL by curtmelonopoly on TradingView.com

Why Liquidity is Critical and Why I Use Instruments like $UWTI (now called $UWT)

Specific to day trading with these algorithms, liquidity is important because you are taking advantage of (exploiting) not only that you know (better than most) what the machines are doing (which provides a distinct edge), but you are more importantly exploiting what other traders are going to do as a result of what the machines have done intra-day (so the goal is to know in advance what the machines will do at various decisions).

In other words, if I know that crude is going to spike because it is near an algo line and I know that when it crosses that line to the upside it will either spike or drop and I know that $REN, for example, is squeezing, then I can exploit that because oil will spike as it crosses the trendline, as will $REN or $UWT.

So my advance knowledge in relation to the probability of that spike enables me to exploit that spike in an equity or ETN that returns unusual short time frame returns.

This page link on our website will show you real life examples of how EPIC the Oil Algo allows me to exploit the algorithmic knowledge I have.

Liquidity gets me predictability for spikes that I need for entries and it also allows me to chip out of large entries when needed in a predictable way.

Where and How Are The Algorithms Available?

We have a main trading room that is like any other trading room where I perform trades during regular market hours. Our algorithmic charting is represented at times in that room but never in whole and only as they are in their initial development phase. Once the algorithm is at a point of proven predictability we then will move it to its own trading room (like EPIC the Oil algo is getting his own 24 hour trading room for crude oil futures).

Also, subscribers to algorithm newsletters get regular updates on that specific algorithm (most are daily but can be intermittent depending on indicators changing). So the subscribers to the specific algorithms are receiving all the algorithmic indicators, trading levels, targets, algo lines etc on a regular basis – subscribers to the main trading room are receiving the benefit (or bonus) of algorithmic charting at various times while an algorithm is in its infancy and being tested or represent in different ways for various reasons.

Algorithm Performance

The performance of the algorithms has completely surprised me – I seriously thought I would be doing great if they achieved a 60% win rate – 90% plus I never fathomed. But what is more interesting (and keep in mind they are only seven months old and each is at a different level of development) is that they are getting more refined and more predictable (with higher win rates) as time goes on. So this is, at least so far, very encouraging.

Plans Going Forward

Right now we are in what we consider a beta phase. We have a main trading room and we have started the process of publishing the six different algo newsletters that represent the algorithmic and traditional charting. Subscribers can use just the room, our swing trading newsletter service or the algo newsletters independent of each other.

This Monday, we publish the remaining charting for all the algos – they are all in a development process at different stages and some of them we haven’t published the charting yet.

Early 2017 we expect EPIC to have his own 24 hour room for oil futures trading and we expect the other five algorithms to be completely proven within a few months. Our oil algo is farthest along in terms of development/proof-of-concept, and our Gold, Silver and US Dollar algos are not far behind oil and are further along in the process than our S&P 500 and VIX algos. Gold, Silver and US Dollar algos should be completely proven out by March 1, and the S&P 500 and VIX by May 1. We are also working on a natural gas algo (as mentioned above), but it isn’t as predictable.

We have others traders coming on in early 2017 to run rooms that focus on trading options, swing trades and momentum plays, as well.

Our future forward plans include a multi-room platform wherein our vision for a democratized environment is developed for Wall Street (as it applies to algorithms being available to the common man – ones that actually work), and we have plans for big data and artificial intelligence (with the goal of increasing our win rates).

Our Guarantees & Pricing Structures

Recently, I had a few traders ask me about our price increases, so I thought I would comment a bit here on that with our reasoning and a Christmas guarantee.

Our trading room is within the typical range at 199.00 a month (for a room that runs charting, screen sharing, algo development, detailed trade broadcasting, a detailed daily premarket newsletter with charting, etc) and with the initial discount code of 38.2% on an annual membership of 990.00/with discount just over 600.00 (or about 50.00) a month – that’s a steal in my thinking.

Our algos currently range in price from 30.00 a month to 500.00 per month depending on how far along they are in development. And even the most expensive, at 500.00 per month, is available at 1,999.00 per year and with a 38.2% discount on the first order is just over 1,200.00, or 100.00 per month for a high performance algorithm.

If a trader can’t return that 100.00 per month or more with EPIC or the 50.00 with the trading room, then they are doing something wrong with the information provided or the information is faulty and we shouldn’t be in this to begin with (the trader’s edge is the whole purpose for doing this). And even at the monthly rates without the one time discount, there should be no reason a trader can’t get a fantastic return on investment. My point? The algorithm or service you are subscribing to and its related cost should be in accordance to the return on investment – I believe without a shadow of a doubt that our pricing achieves that.

The first guarantee we will give you is that the algorithm prices are going up as the algorithms are developed (and they need to as more staff are hired, more office space is needed and more equipment is needed to run the calculations) – but we will guarantee our early adopters the original price paid as long as they continue to subscribe – we don’t care if that is for years – early adopters get the bonus. Late adopters will have to pay fair market value and that’s totally fair and equitable in our thinking. See the terms and conditions on this before subscribing please.

And to our second guarantee, to be absolutely sure we have done everything we can to be sure we stand on this we are prepared to guarantee your investment in our service. In other words, if you sign on we’ll guarantee your first subscription cost 100% – more specifically that you will at least return that amount of profit within the duration of the subscription to cover the cost. There are some conditions, specific to your sharing your trades live and providing documentation (terms and conditions on this guarantee you will be able to find on our website before Dec 24, 2016). So if you agree to the terms we will guarantee that for you. So we’re taking the risk out – that’s the confidence we have in our service.

And finally, if you are a full-time student paying your own way or underprivileged in an extreme manner and don’t mind sharing your story with us and you need a leg up, then send a private DM to me personally on my Twitter and I will consider anything to give back to the community that has been good to me (we may also have some traders that would consider sponsoring you). Either way, we can look at that on a case-by-case basis and I’m not guaranteeing anything because we each only get so many of these “credit codes” to distribute annually. My Twitter handle is @curtmelonopoly. Do me a favor and try and do it before our media interviews start at the end of December 2016.

The link to our subscription shop page is here.

So that’s my post on why our algorithms are different, how we came along to launch a service like this, how I use the algorithms for my trading. pricing structure and our plans going forward.

Any questions at all email us anytime at [email protected].

Best to you and yours!

Curtis

Our algo Twitter feeds:

$WTI (@EPICtheAlgo), $VIX (@VexatiousVIX), $SPY (@FREEDOMtheAlgo), $GLD (@ROSIEtheAlgo), $SLV (@SuperNovaAlgo), $DXY (@DXYUSD_Index).

Article Topics: Compound Trading, Algorithms, Trading, What Makes Our Algos Different, Stocks, Trading, Oil, S&P 500, Silver, US Dollar, $VIX, Volatility, Gold

 

30 thoughts on “Why our Stock Algorithms are Different Than Most

Comments are closed.