Tuesday, 6 November 2018

I love Comice pears because they remind me about successful algo trading

I've written a few times about my introduction to algo trading but Id like to share, perhaps again, how it all started for me.

I was living in Connecticut in the U.S. in the early to mid 1980s. I read an article in the newspaper, probably the New York Times - I miss those newspapers that I loved but never read anymore - about artificial intelligence.

I immediately thought how wonderful it would be to use that for trading. I then came across the name "David Aronson". I can't remember where or how but it was in relation to a new idea called "optimization of trading systems".

I put the two together and contacted David Aronson and set up a meeting in New York City, just a short train ride from where I lived at the time. Cutting short the story of a very interesting meeting with Dave who was yet to become well known, he told me he could optimize any trading system I designed using the indicators I was using in my CompuTrac group then running mostly on Apple II hardware but transitioning to the new IBM Compatible. Sadly, the price of just one optimization was $25,000 as he was using a mini or mainframe computer and had to program my strategy from scratch. I passed.

Luckily, I then came across Bob Pardo who had just releases software for sale that ran on the "new" IBM compatible and could optimize a trading system albeit mostly using moving averages and Wells Wilders new indicators, if I remember correctly.

I went out and bought my IBM Compatible which was a Compaq so=called portable if you were strong enough.I ended up helping Bob run some research. As thanks, he sent me a Xmas basket which was mostly full of Comice pears.

And that's why after a day of successful algo trading I immediately think of Comice pears and want one.

Here's a screen grab of one of the many portfolios of algos I am currently running. The idea is to run as many different algos as possible to smooth out my equity curve. However, its up to you whether you want to spread out your risk in as many baskets as possible or to put all your eggs in one basket and watch that basket very carefully.


As you can see, this portfolio runs eight different algos each trading just one contract. Today's run produced just 3 trades - the other algos didn't find a trade - and there were no losing trades today. The profit for the day was over $1,400. The margin was day trade margin. Stats for these algos showed max losses of about $1500 a trade and win rates between 60% and 75%.

Wednesday, 31 October 2018

You Only Need to be Smart Once

When I started this blog back in 2009 I started talking more about order flow. In those days seeing order flow was more difficult so we used indicators to see order flow's footprints. Then things got better as technology improved and we can now see raw order flow, measure it and analyse it. This was a quantum leap.

During the last 20 years or so the algos have become more pervasive. To many, this has been a bad thing as the algos took away the edge they had, especially the guys in the pits. On the other hand, we electronic locals have thrived if we embraced the so=called new normal.

The next evolution was the Big Bull Market that came out of the crisis of 20072008. Interest rates became zero and stocks became the only game in town.

All change. We now live in an environment of divisiveness
  • USA v China tariffs and more
  • United Kingdom v EEC (Brexit)
  • Zero Interest Rates v Growing Interest Rates
  • Trump v anti-Trump
  • Republican v Democrat (not the same as above)
  • racist v anti-racist
  • and more
The product of all this divisiveness is volatility. We can live or die by the tweet, at lest as far as daily P&L is concerned. The daily and intra-day charts show how this is playing out.

Retail traders/customers have long been the cannon fodder of the Wall Streeters. This has changed for a lot of us as we use technology to put ourselves on an even playing field with "them".

So what has the medicine been?

There are two ways of making money as retail short term traders:
  1. Be very good at reading and trading order flow. Sit at our workstations for a session and be highly focused and take what the market gives us. Be smart every day, every hour. This is doable. Very doable and if you have learned your craft very well, have the right technology and even just a modest amount of capital then there is money to be made, and,
  2. Be smart once and develop an algo that just keeps on giving with a modest amount of gardening.
Wall Street had the same choices. Banks have/had huge trading floors with rows and rows of desks full of traders each trading his bit of the book. Every 8 hours the book was passed forward to the next time zone.

That has changed.

The banks have gone the technology route and so have I, mostly. Quants are the Kings of Wall Street.
I trade as a discretionary trader just on hour or two a day and not every day now. But my fleet of algos trade almost 24 x 6. How do you eat a cow? Just one bite at a time. For example, one of my ES algos just made this one trade yesterday.


That trade made $200 per contract less fees. Now multiply that by 20 or 30 algos or more, each working almost 24 x 6, and on any one day some making money, some losing money and some not trading. If Ive been smart once then the math works for me and I can harvest good profits every month with about an hour and half of effort every day at a time of my choosing.

We can all be electronic locals, just like the big banks.

Sunday, 16 September 2018

Are You Getting Cheesed?

Are You Getting Cheesed? It was Terry Prachett who coined the term "cheesed" in one of his books. "Cheesed" is like getting creamed but its longer and harder.

The feedback I have been getting from traders is that trading is tougher these days. Our job as traders is to identify what is happening by looking at order flow and context. That has not changed. While volatility is continually changing, if we keep to our basic methodology then we should stay on track. If you go right back to the start of this blog now over nine years ago and step forward the you will see that really nothing has changed. However, our tools to look at the markets have improved dramatically.

The chart below is a good example. NQ futures from Friday. Just another day.
 
  Now compare the above chart which clearly shows what "they" are doing and is probably tradeable as it is, with the chart below which has two added elements of context.


The key to understanding what is happening is to relate the order flow to the context. How far are we from the mean? Where are there large volume nodes?  Price and volume without context is not too meaningful.

So if you are getting cheesed, go back to basics. Keep it simple. While trading is not easy only because of how we tend to think, our process can be simple.



Friday, 3 August 2018

The Most Important Indicator in NinjaTrader

There is a very often repeated theme in this blog going back to when it began in October 2009: the math of trading has to work.

If your stats are not right there is no way you can get to CP.

This is an aspect of trading that most people don't focus on at all. It's all about entries for most people and this is fine if that includes knowing where your exits have to be for you to be CP.

When I enter a trade I have a good idea of what the market has to do for me to be wrong so I know what my loss will be. I also know what my win rate is going to be, more or less. Those two numbers tell me what price has to be exceeded for me to be CP.

If my win rate was 34% and my stop was 2 points away then I need a profit of at least 4 points to break even. So I look at my charts to see where resistance is. If its les than 4 points from my proposed entry then the trade is not "good".

If my win rate is 67% and my stop is 2 points away then my break even is 1 point from my proposed entry. I can then need resistance to be only more than 1 point away.

Of course I am looking for trades where resistance is far enough away to make 3, 4 or more times my risk. Reading the order flow and trading the swings from accumulation to distribution allows this to happen often enough.

Well, what's that MOST IMPORTANT INDICATOR in NinjaTrader? Its in the pic.


It's the Risk-Reward tool found under Drawing Tools. Using it can make you CP.

Thursday, 2 August 2018

Love These Trend Days

Price opened below value and it was a a matter of watching when order flow started buying responsivly. That happened almost right off the bat. Big CCI overbought and turned up. In fact, I was long before RTH based on what had happened in the London session.

The volume profiles showed value moving up all day. I took the pic at about 1 pm EST and it was then a question of whether my Fav Fib target would actually be hit or not. By then it didn't matter a lot for me as I was almost out of bullets after scaling out all the way up. These trend days can be recognized by the narrow breadth of TPOs. I start off by buying the order flow and Profile and then make my exit strategy based on what happens next. The Trend Days cannot be predicted after the RTH close but often the Asian and London sessions give a good clue.

I tend to scale at resistance points as those areas may be all that's in the trade. First scale is usually 50%. I am very mindful of the fact that the math has to work: win rate and multiples of risk need to provide a very positive expectancy.



Tuesday, 17 July 2018

Then and Now

We keep talking about how the markets have changed since the pits disappeared but I haven't really said how.

On the floor our job was to provide liquidity. That meant that we faded most moves. I've spoken about the inside out trades and the outside in. Those were with the trend and against the trend. My focus used to be outside in until a trend day appeared.

So what's happened now? Well, the swings have gotten longer most of the time. That means its more profitable to trade inside out - trade with the trend. The HFTs have taken on providing the liquidity so all I need to do is identify the order flow and trade with "them".

The chart I posted in the previous blog and which I repost today is a good example of what I'm talking about. The move starts on the left hand side of the chart and tops very clearly.  Most of the move was the trend up. Now when the order flow changed direction I can go short, not for just a quick scalp, but to trade the retracement swing. The idea is to stay in trades longer - more bars - no matter what periodicity I'm trading.



Tuesday, 10 July 2018

Trading Outside RTH

There are lots of opportunities in the emini ES futures outside of the regular trading hours of the stock market, especially after the RTH close. There are many, many traders and institutions who have to re balance positions after each settlement period. Also, Asian markets are opening for a new day when U.S. markets have just closed.

Its at these times that markets find themselves out of balance on the majority of days. Additionally, I think its easy to find a bias at this time so understanding the order flow is very easy.

The pic below shows the settlement close on the left and the reopen into the "new" day. I use the series of volume profiles to track what is happening in the order flow and the bars for execution. You can see the moving averages that can be used to trigger a trade if you need some extra training wheels or a specific reason to trigger the trade. This I guess is more psychological than tradecraft but if it helps ............................. It is quite simple but not simplistic as there are quite a few nuances to be learned to trade like this.