• ML^2 platform: what can it do

  • Windows in time

  • core structure

  • phase transition

  • CSi – core structure index

  • price entry

  • portfolio monitoring

  • PAIRs

  • Price path

  • Moments to Stress

  • Intra Day tools

ML^2 platform: what can it do

The ML^2 platform contains all the tools and insights needed for any active trader in derivatives to create and achieve their trading objectives

= when we men ALL, we mean ALL the tools

  • no other platform has these tools, nor are they replicable, nor are they available on any other platform

  • you have access to these TOOLs because you have an ML^2 token – you are a member of a small universe of traders with access to the ML intelligence

  • you have work to do, in that the ML^2 intelligence gives you a series of time windows in which it tells you the:

  1. core structure [persistent-momentum / random / mean reverting

  2. duration of that structure

  3. price path for H,L,C, extreme prices in the selected time window

  4. the optimal entry price based on limit orders

  5. the optimal leverage

  6. when the core state will change – to the new price path, ahead in time [2 time units]

So you know whether the price is trending, mean reverting in nature, the forward price path = or where the price will be in time, the time left in the trend, and when the price path – or the new core state will begin ahead in time

Simple statement – hard act to follow
As most trading applications would say the same thing, yet still 95+% of traders still lose their account equity or simply do not meet their trade objectives

  • so that is the difference, the ‘why’ ML^2 is different

ML^2 is divided into multiple modules designed to provide specific information that is needed in any trading program: framework

  • leverage, market risk, portfolio construction, intra-day trading program {discretionary – automated and hybrid version}

The real difference is in the ‘core state programs
[as individual tools in the ML platform]:

  • core structure

  • phase transition

  • price path

  • moments to stress

  • are the critical applications that define ML^2 intelligence

If I were to guess what most active traders are missing: what is needed to enable them to achieve their trading objectives, it is to know where the price of the instrument will be in specific time window – be that the next 1 minute, the next 15 minutes, 30 minutes, 60 minutes or over the next 24 hours

  • Set your leverage so as to maximise leverage, but also to remain in the trade – the not to be stopped out leverage
  • Know the number of time units to margin call – moments to stress = so you can set your stop loss to be outside the price range
  • Set your target price for exit
  • Know the forward price path
  • Know the market price structure = that is if the price path is in a momentum–trending structure or a mean reverting structure
Then know when the market price structure is about to change and the new price structure ahead in time – phase transition
  • the current 15 minute price structure may be ‘persistent momentum LONG’ but will change to ‘mean reverting LONG’ or ‘persistent momentum SHORT’ over the next two time windows = using the 15 minute time unit, the price change will occur over the next 30 minutes

For the active trader [you], there is some work to undertake:

Mainly in working out what best works for them – that is it maybe to simply look at the CSi index in its raw state and viewing the changes to the CSi sequence in order to pick the core state and the phase transition = to seeing the CSi graphics either in itself or in conjunction with the instruments price path – or in tabular format – either numbers or as statements

  • the other tools support risk management with both the exposure and the instrument chosen: leverage, cash buffer, target price / stop price

If you run multiple instruments you can select the portfolio monitoring program, and manage the risks in the instruments that will result in stressing the future $valuation of the portfolio – all in advance through time

  • know the price path = know all that is needed

windows in time

Essentially the ML program works through the creation of time windows for each instrument. The price path is projected forward through time to show the path the instruments price will travel through to the event horizon [= exit price and that time block]

– showing the High, Low, Close, extreme High extreme Low and Worst prices

  • the sequence of time windows that show the best time to exit or enter the trade: the shortest time window will be the 1 minute window which will give you the price path to immediately pick the best time/ price to enter or exit the trade

  • move from the shortest out to longer time windows to see the price path for the whole price trend, determine your trade strategies and monitor the pathway the price takes to the projected price

  • always be aware that the core state can change to a new core state - this is phase transition, so the trader gets advanced warning of imminent price changes

  • the phase transition time window lasts for 2 time units and gives you the advanced warning to action an exit of the trade as the price path will change – sell while the price is still rising, buy while the price is falling but then to rise

So there is a window in time, you are given the signal to buy, or sell the financial derivative – you are given the price path / the price entry point / the nature of the derivatives ‘state’ [persistent, mean reverting and random] = so you can map out your trade from price entry, exit, leverage, stop – even whether you want to trade the derivative – looking for instruments that suits you trade style [prefer mean reverting to price momentum]

As to the mapped price path in the time window – you can start at the daily price window and move to the hourly, 30 minute and 15 minute and then zero in to the 1 minute for more precise entry / exit timing

So if you can picture the series of time blocks in which you action the trade – 1minute through to 24 hours, each with projected prices out to 10 time units = you get to see the geographical 2D landscape of the instruments price path unfold

Within each window you can set the price levels to buy or sell, set stops

Using the optimal CSi algorithm, which sets the optimal time to act [execute the trade] as within 2 time units from the beginning of the phase transition time window –

  1. In the core state that is ‘persistent-momentum’ you would ride the trend to exit [the new phase transition from the current core state to the new core state] – most momentum trends last 6 – 10 time units

= the persistent-momentum trade program would be a series of buy1, sell2, buy2 sell2 sequence in establishing the position: buy1 [establishing the position] then selling 2 [now short1], buying 2 [now long1<]

= the ‘persistent-momentum’ trade program would be a series of buy1, sell2, buy2 sell2 sequence in establishing the position: buy1 [establishing the position] then selling 2 [squaring the long, establishing the short1], buying 2 [squaring the short, establishing the long1]


core state: persistent-momentum LONG

trade sequence: Buy 1 Lot [long]

{phase transition} new core state: persistent-momentum SHORT

trade: Sell 2 Lot [short]

The directional price movement is where tomorrows price will move in the same direction as today’s price moved

  1. In a ‘mean reverting’ core state you will be turning over the trade as the duration of the trend [remember the today’s price is more than likely to reverse direction from yesterdays price direction = if yesterdays price rose, today or tomorrow the price is likely to fall] – we are generally looking at 2 – 3 time units before the price path will change, so the trade is more a buy1 – sell1 [square], buy1 – sell1 [square]: series of trades


core state mean reversion LONG

trade sequence: Buy1 [long] Sell1 [square]

Buy1 [long] Sell1 [square]

The trade style is not to be squaring the exposure in phase transition as the ‘mean reverting’ core state is simply a series of short term directional price movements in wave cycles where the next day’s [time unit] price will likely to be the reverse of today’s price movement

core structure

Core structure details the price path for each instrument extended from its current core state [spot price]

- there are 5 core states

  • persistent momentum LONG

  • persistent momentum SHORT

  • mean reverting LONG

  • mean reverting SHORT

  • random

These descriptors describe the current market price structure for each financial instrument:

  • so a ‘persistent-momentum’ price structure is a trending / momentum price path where tomorrow price will move in the same direction as the current time window directional price movement

  • a ‘mean reverting’ price structure is one where tomorrows price will be the opposite direction to today’s price movement [or the probability that the next time window the directional price movement is in the opposite direction]

Time window is simply the chosen time block – so the 15 minute time window is 15 minute time blocks where we take the spot high, low, close for every 15 minutes time intervals, and then calculate the price path for every next 15 minutes forward for 10 time blocks or in this case 150 minutes [in 10 x 15 minute time lines]

Every instrument is at any point in time in one of the 5 core states – and these core states will change regularly, say every 2 - 3 time units for ‘mean reverting’ and 6 – 10 time units for ‘persistent momentum’ core states:

  • it is when the current core state is about to change that we want to be alerted and to know the new core state we are about to enter

  • so we could be moving from a current core state in ‘persistent momentum LONG’ price structure to where the price is about to fall – or to a core state of ‘persistent momentum SHORT’

  • the change in core state is called phase transition

  • phase transition will also tell you the new core state [ahead in time]

phase transition

Once the core state of the financial instrument is known – then you simply set up the price entry, stop loss, target price and leverage + cash buffer – you know the time lines to exit, you can manage to maximize leverage when the stop price moves outside of the projected price path – then to safely increase the leverage to the event horizon [target price]

  • you know the time and the price path to exit

  • what else do you need to know

To run a continuous trade cycle, as with automated programs, whereby you trade the price trends [waves cycles] the trader will need to monitor the phase transition tool

  • phase transition alerts [forewarns] of an imminent change to the price path, and presents the new price path core state

  • so the trader is given an alert that the current price trajectory will change, even if the price remains moving in its current direction

  • the trader will be given 1 - 2 time block windows alert

  • so you can exit the trade, capture the majority of the trends basis points and prepare for the new state and price entry

The trader should be seeing this in the price path graphs

What you are doing, is riding the price ‘cycles’ or price ‘waves’ and understanding what the expected size and duration of the price ‘wave’ is and when it is about to end [time to get off]

Now you can simply exit the trade and then look for the next core state that you prefer [select the instrument with the current and project core state price path you want to trade – say you want a mean reversion LONG instrument] – that is if you are a momentum trader, you pick instruments currently in, or about to be in persistent momentum [use the Mi watchlist tool to alert you to the style of trade you want]

or the trader can select to exit the trade [zero the exposure] and reenter the trade setup based on the new core state

optimisation of the phase transition time window

There are 5 algorithms used to classify the core state, that use 10 different historic price time periods across which the CSi is calculated:

[5 x 10 = 50] CSI for each instrument to chose from:

The CSi is calculated for the following time blocks [time-windows]:

1 minute, 15 minutes, 30 minutes, 60 minutes and 24 hours [EoD]

= 50 x 5 = 250 CSi for each instrument

  • each CSi will determine the core state, the time duration in that state and the phase transition to the new core state

  • but with different time lags in the phase transition time window

  • that is the optimal CSi chosen gives 1 - 2 time windows advanced warning – so the price of the instrument will remain in its current directional state until 1 - 2 time windows have elapsed, whereby the price direction will change into the new core state [which is also presented]

  • the 2 time window advanced warning sets the trader up to exit the trade at or near the peak/ lowest price and to enter into the new price path

The only input to the CSi is log return – which is only1 of 5 further inputs into the price path

CSi – core structure index

CSi is an index that has been normalized to move between 0.00 and 1.00 and it is both the direction it is moving and the point between 0.00 and 1.00 that determines the instruments core state

So the CSi will dictate what the financial instrument price will do, and when the CSi changes direction [moving upwards, to downwards in direction] then the core state will also change

  • you can see this on the graph for any instrument, in how the price of the financial instrument will follow the direction of the CSi and when the CSi changes direction, then the price of the instrument will follow the new direction of the CSi, but with an optimised time lag of around 2 time units

The trader can simply stay in the trade and its core structure until the advanced alert that the price path will change, and then set up exit and reentry price strategies

There are 5 algorithms – each one suited to a particular instrument style – that is a stock index or ETF will move differently to each individual constituent instrument, and so we use 1 particular algorithm as opposed to another which is better suited to individual shares, or FX, crypto-currencies or commodities

Another algorithm is more suited to mean reverting instruments, where the price path is more prone to reversals to and through the mean price – so higher turnover trading per unit of time with the resultant higher returns than the momentum strategy

price entry

The tool indicates the optimal range in the entry or in the exit prices for the selected instrument in the next time window

= traders will use this tool to set the limit orders [contrarian price strategy] or when to enter using market orders

  • essentially the tool is used to indicate the required entry [exit, or stop] price for each trade

  • a range in prices is given so you can set appropriate limit orders or time market orders

  • timing the trade entry will also help in minimizing the chance of margin call or being stopped out

portfolio monitoring

The monitoring tool portfolio of derivatives / physical financial instruments [3 or more instruments over a longer investment time horizon]

  • the trader-investor who has a number of longer term investments

  • trade them individually, which requires each instrument to be monitored

  • or they can all be monitored in the portfolio monitoring program

So what does the ML portfolio monitoring program do:

  • indicates the core state of each instrument and when that state is in phase transition

  • alerts to the instruments that will cause stress to the portfolio value [= capital loss]

  • indicates the price path for each instrument

  • indicates each instruments forward MTM $value [using closing prices]

  • indicates the portfolio forward MTM $value

  • provides a market risk indicator, that measures the heat in the market = the market price structure volatility and its climatic impact on the portfolio value = getting hot or warm or colder

[the canary in the coal mine]

what to do with this information

As the forward value of the portfolio is mapped, the trader can see when the peak value of the portfolio is achieved [in forward time], the instrument causing the stress and so can instigate capital protection hedges, or rebalancing of the instrument weights [weighted more to cash] or remove the ‘bad apples’ and replace them with new instruments

=> maximise returns over the longer term horizon in a stable-static portfolio – the main attributes would be active hedging of poorer performing instruments at any point in time [while retaining them in the physical portfolio] and rebalancing the instrument weights back to the preferred weights

sleep@night LVR: is the indicator that moves with the change in the market risk of the portfolio [the risk thermometer]

    • defined as the amount of cash buffer needed above the required margin to limit the possibility of margin call

    • the risk indicator will move higher in $ terms when the market risk is increasing = volatility and the directional price path is moving against positive value for the portfolio


More generally known as statistical arbitrage, or a LONG SHORT trading strategy

  • so you BUY 1 financial instrument and SELL another like instrument [basis point equal value]

  • you are trading the change in the basis points or price differential between each instrument

In standard technical analysis, this analysis would use the Bollinger band tool

  • which has you set up the trade by entering the trade when the spot price is at a certain standard deviation from the mean

  • looking for the price then to revert to the mean

ML program does 4 things better

  • the program has you ride the price cycle in a continuous exposure

  • the trade is measured in terms of change in basis points between the 2 selected instruments

  • so if instrument 1 price is going to rise faster than instrument 2, then you will be LONG instrument 1 and SHORT instrument 2

  • phase transition will tell you in advance when instrument 1 price is about to fall faster than instrument2, and so you will be SHORT instrument 1 and LONG instrument 2

  • for example, instrument 1 could be CBA and instrument 2 could be NAB

  • the basis points difference between CBA price and NAB price moves in cyclical waves just like the price of CBA and the price of NAB individually

  • as you are LONG 1 instrument hedged against SHORT a similar instrument the return you have a capital neutral exposure [in physical trades] but a reduced volatility exposure in derivative exposures

  • so the ML methodology is a continuous exposure program that should make you a minimum of 50+% greater return that the standard bollinger band program on each cycle

    As the exposures are moving in the full wave cycle and not simply from the extreme price then only to the mean price

Price path

The price path for each instrument is projected forward for 10 time units across each of the 7 main time windows: High, Low, Close – Worst price and Extreme High and Low price

Worst price tries to capture spikes in prices [outliners that occur all too frequently] but is a good baseline for maximum optimal leverage [lower the leverage to stay in the trade]

Extreme prices cover off the maximum and minimum price movements and are related to the standard High, Low and Close program and would be the benchmark to set target prices in mean reverting market structures

Technical description:

The spot CSi is projected forward from a specific historic CSi anchor point and is 1 input into the price calculation [call it the multiplier factor] – the methodology projects the forward price, which has the core state, trend duration and phase transition inbuilt

The price path is projected in graphical format for all time windows with 20 historical real prices

  • the CSi is calculated for close price for each time window

  • the close price CSi is used for the High, Low, Close, Extreme and Worst prices

  • the price is projected forward for 10 time units, as this is generally around the maximum time line for each wave in a persistent momentum state

  • an advanced version of the price path will use the CSi for the High, Low and Close prices individually

  • this is because often you find the time window when the price for close, high and low are different, as they all don’t occur in the same time window

  • so if you were to be using the high price to signal the peak in price you will need to use the CSi methodology that adopts the high price to calculate the CSi

: complicated but simple

  • this methodology would be able to pick up the changing relationship between the High and Low ranges and so an indicator for volatility clustering

  • or simply better timing of the entry and exit prices = look to buy the instrument using the LOW price CSi and SELL the instrument using the High price CSi

  • or not, simply use the standard version as Close price

Moments to Stress

This tool, as the name implies, informs on the time period to margin call or stop loss [the enemy of active derivative trader but the life jacket for market maker machines: your broker]

It is handy tool to have if you don’t want to run a stop loss and so the baseline price is margin call – which is a ‘real’ price for each exposure

  • so if you are running maximum leverage and no stop, then this tool should be your primary risk control tool and be monitored

we are expected to make continuous improvements to this tool, more in the advanced warning

When we profile traders for their trading attributes, we look to see how each trader fits to a certain ‘risk profile’ in their risk management methodologies

  • so some traders use stop loss prices, take profit prices, limit orders, programmed ratios of stop to take profit price, maximum or limited leverage, or are momentum traders or mean reverting traders

= put this into a series of back tested modulated profiles and you can determine the Probability of Default

  • even with these tools, 90% of traders loss money to market maker machines [B Book or natural hedge books]

  • so if we run an standard profiling analysis across traders we are capable of deriving a probability of default – this probability of default can be used by market makers to manage the risk exposures via hedging the winners, whilst leaving the high probability of default traders un-hedged

  • the moments to stress tool will screw around with this profiling management program

  • moments to stress should ensure the active trader is comfortable with their exposures projected forward path to margin call [hopefully the projected time period to margin call is a long tail away from the targeted price

  • margin call being the enemy of all derivative traders

Intra Day tools

Tools for intraday trading based around the core state, phase transition and price path programs

  • the intraday time windows include 15minute, 30minute and 60 minute and coverage is to 24 hour [EndOfDay]

  • the instrument universe is not limited – so access to all instruments

  • the core state and price path are presented in graphical and numerical-tabular formats

  • at each time window the CSi is updated for new price data and so the price path will change at every time block, and the core state will change on the commencing of phase transition

One can simply watch the price path of each instrument you have an interest with or exposure to, in using the price path rather than the standard price chart, as it keeps you in the picture of where the price is headed rather than where it has been and is now