Need Help Porting Profitable "Honey Badger" Python Script to Quantopian

Back to Community

edited

The 4 Honey Badger Modes:
http://i.imgur.com/aK99fwK.png

[2015-02-16 23:00:00] Start Date: 2011-12-18 04:00:00
[2015-02-16 23:00:00] Day: 1157 Trades: 91 Frequency 12.7
[2015-02-16 23:00:00] USD 3185.0 / 18745956.9 / 588470.1 ROI
[2015-02-16 23:00:00] BTC 1000.0 / 80366.3 / 7936.6 ROI
[2015-02-16 23:00:00] ***************************************
[2015-02-16 23:00:00] Buy and Hold Gain.: 73.2X
[2015-02-16 23:00:00] Asset Gain........: 80.4X
[2015-02-16 23:00:00] Currency Gain.....: 5885.7X

I trade at tradewave.net and I developed the "Honey Badger" which is a very popular, open source, SMA framework for trading cryptocurrency pairs BTCUSD and LTCUSD. Tradewave is relatively new and very limited on backtesting data. I would like to expose my algo to broader data sets for refinement here at Quantopian. I also have several other scripts I'd like to share... but I need some help getting started; namely: If I could get you guys to port/help me port my basic "Honey Badger" from Tradewave-Python to Quantopian-Python, I suspect I'd catch on rather quickly thereafter and be able to share more in your native "python"; without having to stumble through the syntax differences of ordering, calling talib, selecting trading pairs, logging, plotting, etc.

Here's a backtest 2011 to present:
http://pastebin.com/rjN9SnP5

Here's a screen capture:
http://i.imgur.com/rDc8Hz4.png

With a bit of refinement beyond the open source presented here you can get it to trade like this:
http://i.imgur.com/3bQ5G5N.png ...for 2000X asset gain :D

Here's the open source version script, it trades aggregated 12h SMA's. Average trade frequency is once every 2 weeks. Any help is appreciated, let me know if you have any questions. I look forward to learning more about the Quantopian platform!

I appreciate any input... with or without help; goal = open source Honey Badger port to Quantopian and trading the 2000 YHOO bubble by this time next week, because I NEED more backtest data! :D

...and because Quantopian needs the Honey Badger algo ;)

ETA SEE NEXT POST; 95% PORTED ...but I still need some help accessing SMA data.

Here's the script in Tradewave Python:

import talib  
import time

# Run a 3 month backest up to most recent candle in the interval you  
# intend to run LIVE, enter 'KEYS' displayed at end of log below:


KEY1 = 0  
KEY2 = 0  
KEY3 = 0


def initialize():

    #These are my stored objects 

    storage.ratio = n = (43200 / info.interval)  
    storage.overbought = storage.get('overbought', 0)  
    storage.prev_mode = storage.get('prev_mode', 0)  
    storage.trades = storage.get('trades', 0)  
    storage.action = storage.get('action', 0)  
    storage.signal = storage.get('signal', 0)  
    storage.hold = storage.get('hold', 0)  
    storage.mode = storage.get('mode', 0)  
    storage.currency = storage.get('currency', 0)  
    storage.assets = storage.get('assets', 0)  
    storage.start_currency = storage.get('start_currency', 0)  
    storage.start_assets = storage.get('start_assets', 0)


def instrument():

    # This detects the currency pair chosen for backtesting;  
    # bitcoin, litecoin, Euro, and USD are the options  
    # It then accesses my current portfolio balance on each tick  
    # and extracts "USD" "BTC" "LTC" or "EUR" from "info.primary_pair"

    pair = info.primary_pair  
    if pair == pairs.btc_usd:  
        pair = 'btcusd'  
    if pair == pairs.ltc_usd:  
        pair = 'ltcusd'  
    if pair == pairs.ltc_btc:  
        pair = 'ltcbtc'  
    if pair == pairs.btc_eur:  
        pair = 'btceur'  
    if pair == pairs.ltc_eur:  
        pair = 'ltceur'  
    currency_code = ''.join(list((pair)[3:6]))  
    asset_code = ''.join(list((pair)[0:3]))  
    instrument_pair = asset_code + '_' + currency_code  
    storage.instrument = pairs[instrument_pair]  
    storage.currency = portfolio[currencies[currency_code]]  
    storage.assets = portfolio[currencies[asset_code]]  
    storage.currency_CODE = (currency_code).upper()  
    storage.asset_CODE = (asset_code).upper()


def ma12h(timeperiod, depth):

    # This definition calculates SMA via TAlib  
    # for 43200 (12h) candles, regardless of candle size the bot is deployed on  
    # Depth is the number of candles into the past

    sma = data(interval=43200)[info.primary_pair][depth].ma(timeperiod)  
    return sma


def chart():

    # This just plots all the indicators, including  
    # "mode" which is on a secondary Y axis  
    mode = storage.mode  
    if storage.mode < 2:  
        mode = 1

    plot('ma2', storage.ma2)  
    plot('ma30', storage.ma30)  
    plot('floor', storage.floor)  
    plot('moon', storage.moon)  
    plot('ma60', storage.ma60)  
    plot('ma90', storage.ma90)  
    plot('ma150', storage.ma150)  
    plot('resistance', storage.resistance)  
    plot('mode', mode, secondary=True)  
    # 1 = green dragon  
    # 2 = capitulation  
    # 3 = red dragon  
    # 4 = cat bounce

    buy = 0  
    sell = 0  
    if storage.signal == 1:  
        buy = 0.5  
    if storage.signal == -1:  
        sell = -0.5  
    plot('buy_signal', buy, secondary=True)  
    plot('sell_signal', sell, secondary=True)

    if info.tick == 0:  
        plot('offset', -15, secondary=True)


def indicators():  
    # This definition moves each SMA or expression thereof to a stored object

    storage.ma2 = (ma12h(2, 0))  
    storage.ma30 = (ma12h(30, 0))  
    storage.ma55 = (ma12h(55, 0))  
    storage.ma60 = (ma12h(60, 0))  
    storage.ma90 = (ma12h(90, 0))  
    storage.ma150 = (ma12h(150, 0))  
    storage.ma2i = (ma12h(2, -1))  
    storage.ma30i = (ma12h(30, -1))  
    storage.ma55i = (ma12h(55, -1))  
    storage.ma60i = (ma12h(60, -1))  
    storage.ma90i = (ma12h(90, -1))  
    storage.ma150i = (ma12h(150, -1))

    storage.floor = Decimal(0.75) * storage.ma55  
    storage.moon = Decimal(1.05) * storage.ma55  
    storage.resistance = (storage.ma30 + Decimal(2.8) *  
        (storage.ma30 - storage.ma60))

    # Here I define my starting balance

    if info.tick == 0:  
        storage.start_currency = (Decimal(0.00001) +  
            storage.currency + storage.assets * storage.ma2)  
        storage.start_assets = (Decimal(0.00001) +  
            storage.assets + storage.currency / storage.ma2)  
        if storage.currency + storage.assets == 0:  
            log('Zero Starting Portfolio')  
            raise Stop()

   # Here I define whether or not I'm holding assets and currency on each tick

    storage.holding_assets = False  
    storage.holding_currency = False  
    if storage.currency > (storage.ma2 * Decimal(0.12)):  
        storage.holding_currency = True  
    if storage.assets > 0.12:  
        storage.holding_assets = True


def mode_select():

    # Here I define "Green Dragon and Red Dragon"  
    # Cat Bounce always follows Red Dragon  
    # Capitulation always follows Green Dragon

    if storage.hold <= 0:

        ma2 = storage.ma2  
        ma30 = storage.ma30  
        ma60 = storage.ma60  
        ma90 = storage.ma90  
        ma150 = storage.ma150  
        ma30i = storage.ma30i  
        ma60i = storage.ma60i  
        ma90i = storage.ma90i  
        ma150i = storage.ma150i  
        resistance = storage.resistance

        # Green Dragon Market  
        if ((ma30 > ma60) and  
            (ma60 > ma90) and  
            (ma90 > ma150) and  
            (ma2 > ma30) and  
            (ma30 > Decimal(1.002) * ma30i) and  
            (ma60 > Decimal(1.002) * ma60i) and  
                (ma90 > Decimal(1.002) * ma90i)):  
            storage.mode = 1  
            if ma2 > resistance:  
                storage.overbought = 4*43200  
            if storage.overbought > 0:  
                storage.overbought -= info.interval  
                storage.mode = -1

        # Red Dragon Market  
        elif ((resistance < ma150) and  
             (ma2 < ma90) and  
             (ma2 < ma150) and  
             (ma150 < ma150i) and  
                (resistance < ma90)):  
            storage.mode = 3

        # Not Dragon Market  
        else:  
            if storage.prev_mode == (3 or 4):  
                storage.mode = 4  # Cat Bounce  
            if abs(storage.prev_mode) == (1 or 2):  
                storage.mode = 2  # Capitulation


def mode_3():

    # This definition is used to trade Red Dragon

    ma2 = storage.ma2  
    ma2i = storage.ma2i  
    ma30 = storage.ma30  
    signal = 0  
    if ma2 < storage.floor:  
        signal = 1  
    elif ma2 > storage.moon:  
        signal = -1  
    elif ((ma2 < ma2i) and  
            (ma2 < ma30) and  
            (ma2i > ma30)):  
                signal = -1  
    return signal


def think():  
    # This is the primary trading logic

    action = 0  
    signal = 0  
    mode_select()  
    n = storage.ratio  
    ma2 = storage.ma2  
    hold = storage.hold  
    mode = storage.mode  
    ma30 = storage.ma30  
    ma60 = storage.ma60  
    ma90 = storage.ma90  
    ma90i = storage.ma90i  
    ma150 = storage.ma150  
    prev_mode = storage.prev_mode  
    holding_assets = storage.holding_assets  
    holding_currency = storage.holding_currency

    if hold <= 0:

        # MODE <2 - GREEN DRAGON  
        if mode == 0:  
            signal = 1  
            hold = 1  
            if holding_currency:  
                log('CRYPTO LONG')  
                action = 1  
                hold = 1  
        if mode == 1:  
            signal = 1  
            hold = 1  
            if holding_currency:  
                log('GREEN DRAGON')  
                action = 1  
                hold = 1  
        if mode == -1:  
            signal = -1  
            hold = 0  
            if holding_assets:  
                log('OVERBOUGHT')  
                action = -1  
                hold = 1

        # MODE 2 - CAPITULATION  
        if ((mode == 2) and (prev_mode != 2)):  
            signal = -1  
            if holding_assets:  
                log('CAPITULATION')  
                action = -1  
                hold = 1  
        if mode == prev_mode == 2:  
            if ma2 < ma90:  
                signal = 1  
                # hold = 0  
                if holding_currency:  
                    log('DESPAIR')  
                    action = 1  
                    hold = 1  
            if ((ma2 > Decimal(1.1) * ma90) and  
                    (ma2 > ma60)):  
                signal = -1  
                # hold = 0  
                if holding_assets:  
                    log('FINAL RUN OF THE BULLS - B')  
                    action = -1  
                    hold = 2

        # MODE 3 - RED DRAGON  
        if ((mode == 3) and (prev_mode != 3)):  
            signal = -1  
            hold = 0  
            if holding_assets:  
                log('RED DRAGON')  
                action = -1  
                hold = 5  
        if mode == prev_mode == 3:  
            signal_3 = mode_3()  
            if signal_3 == 1:  
                signal = 1  
                hold = 5  
                if holding_currency:  
                    log('PURGATORY')  
                    action = 1  
                    hold = 5  
            elif signal_3 == -1:  
                signal = -1  
                hold = 0  
                if holding_assets:  
                    log('DEEPER')  
                    action = -1  
                    hold = 1

        # MODE 4 - CAT BOUNCE  
        if ((mode == 4) and (prev_mode != 4)):  
            signal = 1  
            hold = 0  
            if holding_currency:  
                log('HONEY')  
                action = 1  
                hold = 2  
        if mode == prev_mode == 4:  
            if ((ma2 < Decimal(1.05) * ma30) and  
                (ma2 > ma60) and  
                    (ma90 > Decimal(1.003) * ma90i)):  
                signal = -1  
                hold = 1  
                if holding_assets:  
                    log('KARMA')  
                    action = -1  
                    hold = 1  
            if ((ma2 < ma90) and  
                (ma90 < ma60) and  
                (ma90 < ma30) and  
                    (ma90 > ma90i)):  
                signal = 1  
                hold = 0  
                if holding_currency:  
                    log('GOLD')  
                    action = 1  
                    hold = 1  
        storage.hold = Decimal(hold) * 86400

    storage.hold -= info.interval  
    storage.signal = signal  
    storage.action = action  
    storage.prev_mode = storage.mode


def order():  
    # Here is how I actually place orders at Tradewave  
    if storage.action == 1:  
        if storage.holding_currency:  
            try:  
                order = buy(info.primary_pair)  
                storage.trades += 1  
                email(order.price,  
                    subject='**ALERT** HONEY BADGER SAYS: BUY OR CRY')  
            except TradewaveFundsError as e:  
                log('Not enough currency!')  
            except TradewaveInvalidOrderError as e:  
                log('Unable to place buy order: %s' % e.message)  
    if storage.action == -1:  
        if storage.holding_assets:  
            try:  
                order = sell(info.primary_pair)  
                storage.trades += 1  
                email(order.price,  
                    subject='**ALERT** HONEY BADGER SAYS: FIAT TIME')  
            except TradewaveFundsError as e:  
                log('Not enough assets!')  
            except TradewaveInvalidOrderError as e:  
                log('Unable to place buy order: %s' % e.message)  
    if abs(storage.action) == 1:  
        report()


def report():

    # This is used to periodically report ROI via the log

    days = (info.current_time - info.begin) / 86400  
    trades = storage.get('trades', 0)  
    if trades == 0:  
        frequency = 0  
    else:  
        frequency = Decimal(days) / Decimal(trades)  
    start_date = time.strftime(  
        '%Y-%m-%d %H:%M:%S', time.localtime(info.begin))  
    potential_assets = storage.assets + storage.currency / storage.ma2  
    potential_currency = storage.currency + storage.assets * storage.ma2  
    start_assets = storage.start_assets  
    start_currency = storage.start_currency  
    ROI_currency = (100 * potential_currency / start_currency) - 100  
    ROI_assets = (100 * potential_assets / start_assets) - 100

    log('Start Date: %s' % start_date)  
    log('Day: %s Trades: %s Frequency %.1f' % (days, trades, frequency))  
    log('%s %.1f / %.1f / %.1f ROI' %  
        (storage.currency_CODE, start_currency,  
         potential_currency, ROI_currency))  
    log('%s %.1f / %.1f / %.1f ROI' %  
        (storage.asset_CODE, start_assets, potential_assets, ROI_assets))  
    log('***************************************')


def tick():  


    # Tick is called on every tick and directs the other definitions in the script

    if info.tick == 0:  
        storage.hold = KEY1  
        storage.mode = storage.prev_mode = KEY2  
        storage.overbought = KEY3  


    instrument()  
    indicators()  
    chart()  
    think()

    if info.tick > 0:  
        order()

    if 1:  # on/off COUNT DAYS  
        n = storage.ratio  
        if info.tick % (2 * n) == 0:  
            log(info.tick / (2 * n))  
    if storage.hold < 0:  
        storage.hold = 0


def stop():

    # This is the final report that displays in the log during backtesting on the last tick

    bnh_gain = ((storage.start_assets * storage.ma2) /  
        storage.start_currency)  
    currency_gain = ((storage.currency + storage.assets * storage.ma2) /  
        storage.start_currency)  
    asset_gain = currency_gain / bnh_gain  
    if storage.assets*storage.ma2 > storage.currency:  
        holding = storage.asset_CODE  
    else:  
        holding = storage.currency_CODE

    log('***************************************')  
    log('***************************************')  
    log('Honey Badger v1.0 80X by litepresence')  
    log('')  
    report()  
    log('Buy and Hold Gain.: %.1fX' % bnh_gain)  
    log('Asset Gain........: %.1fX' % asset_gain)  
    log('Currency Gain.....: %.1fX' % currency_gain)  
    log('**** INITIALIZATION KEYS ****')  
    log('KEY1 = %s' % int(storage.hold))  
    log('KEY2 = %s' % storage.mode)  
    log('KEY3 = %s' % storage.overbought)  
    log('HOLDING %s' % holding)

Thanks for reading! I appreciate your help getting acquainted to Quantopian!

Here are some other noteworthy scripts I've written in Tradewave-Python you might find interesting:

Iceberg Limit Orders:
https://discuss.tradewave.net/t/iceberg-limit-orders-v1-0-beta-litepresence

Parabolic SAR Custom without TAlib:
https://discuss.tradewave.net/t/parabolic-sar-custom-without-ta-lib

11 responses

litepresence

HONEY BADGER ALGORITHM IN PLAIN ENGLISH

VARIABLES:

12h candle moving averages:

ma2  
ma30  
ma55  
ma60  
ma90  
ma150 

expressions of those moving averages:

floor = 0.75*ma55  
moon = 1.05*ma55  
resistance = ma30 + 2.8*(ma30 - ma60)


DEFINE MARKET:

It is a bull (green dragon) market if:

ma30 > ma60 and  
ma60 > ma90 and  
ma90 > ma150 and  
ma2 > ma30 and  
ma30 has positive slope  
ma60 has positive slope  
ma90 has positive slope

It is a bear (red dragon) market if:

resistance < ma150 and  
ma2 < ma90 and  
ma2 < ma150 and  
ma150 slope is negative and  
resistance < ma90

Its a capitulating market if:

It was previously a bull market and  
It is currently neither bull or bear

It is cat bouncing market if:

It was previously a bear market and  
It is currently neither bull or bear

It is an overbought bull market if:

ma2 > resistance

Bear market fell through floor if:

ma2 < floor

Bear market is continuing if:

ma2 has negative slope  
ma2 crosses under ma30

or if ma2 is over the moon

Despair is when market is capitulating but:

ma2 is less than ma90

Final run of bulls is when market is capitulating, you bought despair and:

ma2 > ma60 and  
ma2 > 1.1*ma90

Bad Karma is when mode is cat bounce, but:

ma2 < 1.05*ma30 and  
ma2 > ma60 and  
ma90 has strong positive slope

Gold is when mode is cat bounce, but:

ma2 < ma90  
ma90 < ma60  
ma90 < ma30  
ma90 has positive slope


TRADE:

>>>>>>>>>>>>>>>>>  
BULL  
>>>>>>>>>>>>>>>>>  
GREEN DRAGON:  
Always buy the beginning of bull market

OVERBOUGHT  
Always sell an overbought bull market

>>>>>>>>>>>>>>>>>  
CAPITULATION  
>>>>>>>>>>>>>>>>>  
CAPITULATION:  
Always sell beginning of capitulation

DESPAIR  
Always buy despair

FINAL RUN OF BULLS  
If holding despair dump final run of bulls

>>>>>>>>>>>>>>>>>  
BEAR  
>>>>>>>>>>>>>>>>>  
RED DRAGON:  
Always dump the beginning of a bear market

PURGATORY:  
If bear market fell through floor, buy it

DEEPER:  
If holding purgatory and bear market is continuing dump

>>>>>>>>>>>>>>>>>  
CAT BOUNCE  
>>>>>>>>>>>>>>>>>  
HONEY:  
Always buy beginning of cat bounce

KARMA:  
Always sell bad karma

GOLD:  
Always buy gold

Tartarus Group

Nice to see somebody else from cryptotrader. I'll give it a look when I get some free time.

litepresence

99% Functional Quantopian Port

Its compiling and trading... but something tells me this isn't backtesting correctly:

Total Returns
251956424.9%
From 2003-01-03 to 2015-02-01 with $1,000 initial capital (daily data)

My goal is to buy all I can afford on all buy events (action = 1), and sell all shares on every sell (action = -1)

import talib

SECURITY = symbol('YHOO')

def initialize(context):  
    context.overbought =0  
    context.hold = 0  
    context.mode = 0  
    context.prev_mode = 0  
    context.action = 0  
    context.signal = 0   


def ma12h(period, depth, context):  
    close = history(160, '1d', 'close_price')  
    sma = close.apply(talib.MA, timeperiod=period).iloc[depth]  
    return sma

def chart(context):

    record('ma2', context.ma2)  
    record('ma30', context.ma30)  
    record('ma60', context.ma60)  
    record('ma90', context.ma90)  
    record('ma150', context.ma150)  
def indicators(context):

    # CURRENT DAILY SMA's  
    context.ma2 = float(ma12h(2, -1, context))  
    context.ma30 = float(ma12h(30, -1, context))  
    context.ma55 = float(ma12h(55, -1, context))  
    context.ma60 = float(ma12h(60, -1, context))  
    context.ma90 = float(ma12h(90, -1, context))  
    context.ma150 = float(ma12h(150, -1, context))  
    # PREVIOUS DAILY SMA's  
    context.ma2i = float(ma12h(2, -2, context))  
    context.ma30i = float(ma12h(30, -2,context))  
    context.ma55i = float(ma12h(55, -2, context))  
    context.ma60i = float(ma12h(60, -2, context))  
    context.ma90i = float(ma12h(90, -2, context))  
    context.ma150i = float(ma12h(150, -2, context))

    # EXPRESSIONS  
    context.floor = float(0.75) * float(context.ma55)  
    context.moon = float(1.05) * float(context.ma55)  
    context.resistance = (float(context.ma30) + float(2.8) *  
        (float(context.ma30) - float(context.ma60)))

    # ASSESS PORTFOLIO  
    context.holding_assets = context.holding_currency = False  
    if context.portfolio.cash > 10*context.ma2:  
        context.holding_currency = True  
    if context.portfolio.positions[SECURITY] > 10:  
        context.holding_assets = True  


def mode_select(context):  
    ma2 = context.ma2  
    ma30 = context.ma30  
    ma60 = context.ma60  
    ma90 = context.ma90  
    ma150 = context.ma150  
    ma30i = context.ma30i  
    ma60i = context.ma60i  
    ma90i = context.ma90i  
    ma150i = context.ma150i  
    resistance = context.resistance

    # Green Dragon Market  
    if ((ma30 > ma60) and  
        (ma60 > ma90) and  
        (ma90 > ma150) and  
        (ma2 > ma30) and  
        (ma30 > float(1.002) * ma30i) and  
        (ma60 > float(1.002) * ma60i) and  
           (ma90 > float(1.002) * ma90i)):  
        context.mode = 1  
        if ma2 > resistance:  
            context.mode = -1

    # Red Dragon Market  
    elif ((resistance < ma150) and  
         (ma2 < ma90) and  
         (ma2 < ma150) and  
         (ma150 < ma150i) and  
            (resistance < ma90)):  
        context.mode = 3

    # Not Dragon Market  
    else:  
        if context.prev_mode == (3 or 4):  
            context.mode = 4  # Cat Bounce  
        if abs(context.prev_mode) == (1 or 2):  
            context.mode = 2  # Capitulation

def mode_3(context):

    ma2 = context.ma2  
    ma2i = context.ma2i  
    ma30 = context.ma30  
    signal = 0  
    if ma2 < context.floor:  
        signal = 1  
    elif ma2 > context.moon:  
        signal = -1  
    elif ((ma2 < ma2i) and  
            (ma2 < ma30) and  
            (ma2i > ma30)):  
                signal = -1  
    return signal

def think(context):  
    holding_assets = context.holding_assets  
    holding_currency = context.holding_currency        

    action = 0  
    mode_select(context)  
    ma2 = context.ma2

    ma30 = context.ma30  
    ma60 = context.ma60  
    ma90 = context.ma90  
    ma90i = context.ma90i  
    prev_mode = context.prev_mode  
    mode = context.mode  
    if 1:

        # MODE <2 - GREEN DRAGON  
        if mode == 0:  
            hold = 1  
            if holding_currency:  
                log.info('CRYPTO LONG')  
                action = 1  
                hold = 1  
        if mode == 1:  
            hold = 1  
            if holding_currency:  
                log.info('GREEN DRAGON')  
                action = 1  
                hold = 1  
        if mode == -1:  
            hold = 0  
            if holding_assets:  
                log.info('OVERBOUGHT')  
                action = -1  
                hold = 1

        # MODE 2 - CAPITULATION  
        if ((mode == 2) and (prev_mode != 2)):  
            if holding_assets:  
                log.info('CAPITULATION')  
                action = -1  
                hold = 1  
        if mode == prev_mode == 2:  
            if ma2 < ma90:  
                # hold = 0  
                if holding_currency:  
                    log.info('DESPAIR')  
                    action = 1  
                    hold = 1  
            if ((ma2 > float(1.1) * ma90) and  
                    (ma2 > ma60)):  
                # hold = 0  
                if holding_assets:  
                    log.info('FINAL RUN OF THE BULLS - B')  
                    action = -1  
                    hold = 2

        # MODE 3 - RED DRAGON  
        if ((mode == 3) and (prev_mode != 3)):  
            hold = 0  
            if holding_assets:  
                log.info('RED DRAGON')  
                action = -1  
                hold = 5  
        if mode == prev_mode == 3:  
            signal_3 = mode_3(context)  
            if signal_3 == 1:  
                hold = 5  
                if holding_currency:  
                    log.info('PURGATORY')  
                    action = 1  
                    hold = 5  
            elif signal_3 == -1:  
                hold = 0  
                if holding_assets:  
                    log.info('DEEPER')  
                    action = -1  
                    hold = 1

        # MODE 4 - CAT BOUNCE  
        if ((mode == 4) and (prev_mode != 4)):  
            hold = 0  
            if holding_currency:  
                log.info('HONEY')  
                action = 1  
                hold = 2  
        if mode == prev_mode == 4:  
            if ((ma2 < float(1.05) * ma30) and  
                (ma2 > ma60) and  
                    (ma90 > float(1.003) * ma90i)):  
                hold = 1  
                if holding_assets:  
                    log.info('KARMA')  
                    action = -1  
                    hold = 1  
            if ((ma2 < ma90) and  
                (ma90 < ma60) and  
                (ma90 < ma30) and  
                    (ma90 > ma90i)):  
                hold = 0  
                if holding_currency:  
                    log.info('GOLD')  
                    action = 1  
                    hold = 1

    context.action = action  
    context.prev_mode = context.mode

def trade(context):  
    if abs(context.action) ==1:  
        log.info(context.mode)    

    if context.action == 1:  
        if context.holding_currency:  
            order_percent(SECURITY, 1)  
    if context.action == -1:  
        if context.holding_assets:  
            order_percent(SECURITY, -1)  




def handle_data(context,data):  
    indicators(context)  
    chart(context)  
    think(context)  
    trade(context)

Tartarus Group

I'd use order_target_percent instead, if you graph leverage you'd notice, you'd see you're over leveraging yourself quite a bit at a few points.

Peter Bakker

when I use order_target_percent as Tartarus suggests the algo performs -74%. However when I turn around the -1 and the +1 in the buying section it is a positive algorithm. With a high DD still.

Peter

litepresence

thanks guys, very helpful

Any thoughts on the 12h moving average in cryptocurrency vs 1d EOD moving averages in stock and how they scale relatively?

BIIB

AMZN

AAPL

hmm. although it performs well on those stocks, I still don't think its trading "correctly"; its basically buy and hold with only a few moves. And these are kind of cherry picked top performers.

Tartarus Group

Stocks generally don't have the volatility that crypto currencies do.

You've successfully submitted a support ticket.

Our support team will be in touch soon.