thetagang.toml

# NOTE: It is STRONGLY recommended you read through all notes, config options,
# and documentation before proceeding. Be sure to update the configuration
# values according to your preferences. Additionally, any default values in
# this config do not constitute a recommendation or endorsement, or any provide
# claims abount returns or performance.
#
# Should you decide to use ThetaGang, please experiment with a paper trading
# account before trying on a live account.
[account]
# The account number to operate on
number = "DU1234567"

# Cancel any existing orders for the symbols configured at startup
cancel_orders = true

# Maximum amount of margin to use, as a ratio of net liquidation. IB lets
# you use varying amounts of margin, depending on the assets. To use up to 4x
# margin, set this to 4. It's recommended you always leave some additional
# cushion. IB will start to close positions if you go below certain thresholds
# of available margin in your account.
#
# For details on margin usage, see:
#   https://www.interactivebrokers.com/en/index.php?f=24176
#
# The default value uses 50% of your available net liquidation value
# (i.e., half of your funds). Set this to 1.0 to use all your funds,
# or 1.5 to use 150% (which may incur margin interest charges).
#
# In other words, ThetaGang's buying power is calculated by taking your NLV and
# multiplying it by margin_usage.
margin_usage = 0.5

# Market data type (see
# https://interactivebrokers.github.io/tws-api/market_data_type.html)
market_data_type = 1

[constants]
# Refer to `durationString` at
# https://interactivebrokers.github.io/tws-api/historical_bars.html for details.
# Valid values include "30 D", "60 D", "6 M", "1 Y", etc.
daily_stddev_window = "30 D"

# Optionally, specify a write threshold either as percent, or sigma, either for
# all contracts, or separately for puts/calls. Refer to the comments for
# `write_threshold` and `write_threshold_sigma` under `symbols` for details.
#
# # Applies to both puts and calls
# write_threshold = 0.01 # 1%
# write_threshold_sigma = 1.0 # 1𝜎
# [puts] # Applies only to puts
# write_threshold = 0.01 # 1%
# write_threshold_sigma = 1.0 # 1𝜎
# [calls] # Applies only to calls
# write_threshold = 0.01 # 1%
# write_threshold_sigma = 1.0 # 1𝜎

[orders]
# The exchange to route orders to. Can be overridden if desired. This is also
# used for fetching tickers/prices.
exchange = "SMART"

# Range of time to delay, in seconds, before resubmitting an order with updated
# midpoint price if `symbol.<symbol>.adjust_price_after_delay = true`.
price_update_delay = [30, 60]

# Set a minimum credit order price, to avoid orders where the credit (or debit)
# is so low that it doesn't even cover broker commission. We default to $0.05,
# but you can set this to 0.0 (or comment it out) if you want to permit any
# order price. This doesn't apply to debit orders.
minimum_credit = 0.05

  [orders.algo]
  # By default we use adaptive orders with patient priority which gives reasonable
  # results. You can also experiment with TWAP or other options, however the
  # available order algos vary depending on what you trade.
  #
  # Note that the algo orders don't seem to work with combo orders, which are used
  # when rolling positions, so AFAIK this has no effect for those orders. It only
  # seems to take effect with regular open/close orders.
  # Optional IBKR algo strategy. See
  # https://interactivebrokers.github.io/tws-api/ibalgos.html for option details.
  strategy = "Adaptive"

  # For `algoParams`, the TagValue parameter has 2 values, so any values with
  # anything other than 2 parameters are invalid. Pass an empty list to use the
  # defaults (i.e., params = []).
  params = [
    [
      "adaptivePriority",
      "Patient",
    ],
  ]

[option_chains]
# The option chains are lazy loaded, and before you can determine the greeks
# (delta) or prices, you need to scan the chains. The settings here tell
# thetagang how many contracts to load. Don't make these values too high, as
# they will cause the chain scanning process to take too long, and it may fail.
#
# If you have issues where thetagang can't find suitable contracts, try
# increasing these values slightly.
#
# Number of expirations to load from option chains
expirations = 4

# Number of strikes to load from option chains
strikes = 15

[roll_when]
# Roll when P&L reaches 90%
pnl = 0.9
# Or, roll options when there are <= 15 days to expiry and P&L is at least
# min_pnl (min_pnl defaults to 0)
#
# NOTE: For cases where an option ends up deep ITM, notably when selling
# covered calls, it's possible that the P&L would be significantly negative,
# i.e., -100%. If you want to roll anyway in these situations, set min_pnl to a
# negative value such as -1 (for -100%).
dte     = 15
min_pnl = 0.0

# Optional: Don't roll contracts when the current DTE is greater than this
# number of days. This helps avoid cases where you end up rolling out to LEAPs.
#
# max_dte = 180

# Optional: Create a closing order when the P&L reaches this threshold. This
# overrides the other parameters, i.e., it ignores DTE and everything else.
# If not specified, it has no effect. This can handle the case where you have
# long-dated options that have slowly become worthless and you just want to get
# them out of your portfolio. This only applies to short contract positions,
# long positions are ignored.
#
# close_at_pnl = 0.99

# Optional: if we try to roll the position and it fails, just close it (but only
# if the position currently has positive P&L). This can happen if the underlying
# moves too much and there are no suitable contracts. See
# https://github.com/brndnmtthws/thetagang/issues/347 and
# https://github.com/brndnmtthws/thetagang/issues/439 for a discussion on this.
#
# If `roll_when.max_dte` is set, this will only close the position if the DTE is
# <= `roll_when.max_dte`.
#
# This can also be set per-symbol, with
# `symbols.<symbol>.close_if_unable_to_roll`.
close_if_unable_to_roll = false

  [roll_when.calls]
  # Roll calls to the next expiration even if they're in the money. Defaults to
  # true if not specified.
  itm = true

  # Always roll calls (short-circuit) when they're in the money, regardless of
  # the P&L. This allows you to avoid assignment risk by eating the cost of
  # rolling the calls. This is useful if you want to avoid assignment risk
  # either because you don't own the underlying (i.e., with spreads) or because
  # you don't want to realize a gain/loss yet. Note that if it's rolling early
  # (i.e., before `roll_when.dte`) it may only roll a subset of the contracts
  # based on the `target.maximum_new_contracts_percent` value.
  #
  # This can also be set for puts with `roll_when.puts.always_when_itm`. This
  # option, when enabled, takes precedence over `roll_when.pnl` and
  # `roll_when.min_pnl`.
  always_when_itm = false

  # Only roll when there's a suitable contract available that will result in a
  # credit. Enabling this may result in the target delta value being ignored in
  # circumstances where we can't find a contract that will result in both a
  # credit _and_ satisfying the target delta (i.e., having a credit takes
  # precedence).
  credit_only = false

  # If set to false, calls will not be rolled if there are any number of calls in
  # excess of the target call quantity. A truthy value means thetagang will keep
  # rolling calls regardless of the total quantity.
  has_excess = true

  # We can optionally maintain the high water mark for covered calls by never
  # rolling CCs down, only out. We do this because it's not uncommon for a sharp
  # bounce up after a stock/index falls enough to make the contract eligible for
  # rolling. If we roll the CC down and out, we might be capping our gains when it
  # eventually does bounce back. The trade off here is that you end up with
  # slightly less downside protection.
  #
  # This only applies to rolling calls, and is essentially equivalent to setting
  # the minimum strike of the next contract to the previous strike and only
  # allowing the value to ratchet upwards.
  #
  # While this is off by default, it would be sensible to turn it on if you're
  # trying to capture upside, rather than protect downside.
  #
  # This value can also be set per-symbol, with
  # `symbols.<symbol>.calls.maintain_high_water_mark` (see `symbols.QQQ` example
  # below).
  maintain_high_water_mark = false

  [roll_when.puts]
  # Roll puts if they're in the money. Defaults to false if not specified.
  itm = false

  # See comments above for `roll_when.calls.always_when_itm` for details on this
  # option (which behaves the same for puts as with calls).
  always_when_itm = false

  # Only roll when there's a suitable contract available that will result in a
  # credit. Enabling this may result in the target delta value being ignored in
  # circumstances where we can't find a contract that will result in both a
  # credit _and_ satisfying the target delta (i.e., having a credit takes
  # precedence).
  credit_only = false

  # If set to false, puts will not be rolled if there are any number of puts in
  # excess of the target put quantity. A truthy value means thetagang will keep
  # rolling puts regardless of the total quantity.
  has_excess = true

[write_when]
# By default, we ignore long option positions when calculating whether we can
# write new contracts. This, however, limits the ability to utilize spreads. For
# example, we might want to buy some long-dated puts when they're cheap, and
# then we can sell additional short-dated puts (i.e., a calendar spread).
# ThetaGang won't open a calendar spread for you by purchasing longer-dated
# contracts, but if you were to buy some LEAPs yourself, ThetaGang could take
# advantage of those positions when calculating whether to write additional
# contracts.
#
# If you set this value to true, ThetaGang will calculate the net positions by
# including the long contracts in its calculations. It does this by greedily
# matching short positions with long positions such that they cancel each other
# out. ThetaGang will only match on those positions where the long leg has a
# greater DTE, and the strike is closer to the money (i.e., the strike for puts
# must be >=, and the strike for calls must be <=).
#
# IMPORTANT NOTE: ThetaGang doesn't manage long contracts at all, so you
# introduce the risk of having excess short positions when the long contracts
# eventually expire. For the time being, you need to manage the long legs
# yourself (by either closing the short side when the long positions expire, or
# rolling the long positions as they approach expiration). You can avoid rolling
# excess positions with `roll_when.calls/puts.has_excess = false`.
calculate_net_contracts = false

  [write_when.calls]
  # Optionally, only write calls when the underlying is green
  green = true
  red   = false

  # With covered calls, we can cap the number of calls to write by this factor. At
  # 1.0, we write covered calls on 100% of our positions. At 0.5, we'd only write
  # on 50% of our positions. This value must be between 1 and 0 inclusive.
  #
  # This can also be set per-symbol with
  # `symbols.<symbol>.calls.cap_factor`.
  cap_factor = 1.0

  # We may want to leave some percentage of our underlying stock perpetually
  # uncovered so we don't miss out on big upside movements. For example, if our
  # target number of shares for a given symbol is 1,000, we may want to ensure
  # 500 shares are _always_ uncovered. This is a bit different from `cap_factor`,
  # as it applies to the _target_ number of shares, as opposed to the current
  # positions. A value of 0.5 (50%) means that we always leave 50% (half) of the
  # target shares uncovered. A bigger number (up to 1.0, 100%) is more bullish,
  # and a smaller number (down to 0.0, 0%) is more bearish (i.e., cover all
  # positions with calls). This essentially sets a floor on the number of shares
  # we try to hold on to in order to avoid missing out on potential upside.
  #
  # This can also be set per-symbol with
  # `symbols.<symbol>.calls.cap_target_floor`.
  cap_target_floor = 0.0

  # If set to true, calls are only written on the underlying when the underlying
  # has an excess of shares. This is useful for covered calls, where you only
  # want to write calls when you have more shares than you want to hold
  # long-term. It also provides a way to rebalance your portfolio by writing
  # calls and taking profits.
  #
  # This may also be set per-symbol with `symbols.<symbol>.calls.excess_only`,
  # which takes precedence.
  #
  # When this is set to true, the `cap_factor` and `cap_target_floor` values are
  # ignored.
  excess_only = false

  [write_when.puts]
  # Optionally, only write puts when the underlying is red
  green = false
  red   = true

[target]
# Target 45 or more days to expiry
# This value can also be set per-symbol, with `symbols.<symbol>.dte`
# (see `symbols.QQQ` example below).
dte = 45

# Optionally, we can prevent contracts from being rolled insanely far out with
# `target.max_dte`. Comment this setting out if you don't care about it. You can
# also specify this per-symbol, with `symbols.<symbol>.max_dte`, or for VIX if
# you use the VIX call hedging.
max_dte = 180

# Target delta of 0.3 or less. Defaults to 0.3 if not specified.
delta = 0.3

# When writing new contracts (either covered calls or naked puts), or rolling
# before `roll_when.dte` is reached, never write more than this amount of
# contracts at once. This can be useful to avoid bunching by spreading contract
# placement out over time (and possibly expirations) in order to protect
# yourself from large swings. This value does not affect rolling existing
# contracts to the next expiration. This value is expressed as a percentage of
# buying power based on the market price of the underlying ticker, as a range
# from [0.0-1.0].
#
# Once the `roll_when.dte` date is reached, all the remaining positions are
# rolled regardless of the current position quantity.
#
# Defaults to 5% of buying power. Set this to 1.0 to effectively disable the
# limit.
maximum_new_contracts_percent = 0.05

# Minimum amount of open interest for a contract to qualify
minimum_open_interest = 10

# Optional: specify delta separately for puts/calls. Takes precedent over
# target.delta.
#
#  [target.puts]
#  delta = 0.5
#  [target.calls]
#  delta = 0.3
[symbols]
  # NOTE: Please change these symbols and weights according to your preferences.
  # These are provided only as an example for the purpose of configuration. These
  # values were chosen as sane values should someone decide to run this code
  # without changes, however it is in no way a recommendation or endorsement.
  #
  # You can specify the weight either as a percentage of your buying power (which
  # is calculated as your NLV * account.margin_usage), or in terms of parts. Parts
  # are summed from all symbols, then the weight is calculated by dividing the
  # parts by the total parts.
  #
  # You should try to choose ETFs or stocks that:
  #
  #  1) Have sufficient trading volume for the underlying
  #  2) Have standard options contracts (100 shares per contract)
  #  3) Have options with sufficient open interest and trading volume
  #
  # The target delta may also be specified per-symbol, and takes precedence over
  # `target.delta` or `target.puts/calls.delta`. You can specify a value for the
  # symbol, or override individually for puts/calls.
  [symbols.SPY]
  weight = 0.4

  # OR: specify in terms of parts. Must use either weight or parts, but cannot mix
  # both.
  # parts = 40

  # Sometimes, particularly for stocks/ETFs with limited liquidity, the spreads
  # are too wide to get an order filled at the midpoint on the first attempt. For
  # those, you can try setting this to `true`, and thetagang will wait a random
  # amount of time, then resubmit orders that haven't filled (but only for the
  # symbols with this set to true). The amount of time we'll wait is chosen
  # randomly from the range defined by `orders.price_update_delay`.
  adjust_price_after_delay = false

  # You can include a symbol, but instruct ThetaGang not to place any trades for
  # that symbol by setting `no_trading = true`. This allows you to include
  # placeholders (such as for tracking purposes, or because you want to trade
  # manually) without actually trading them.
  #
  # no_trading = true

  [symbols.QQQ]
  weight = 0.3
  # The target DTE may also be specified per-symbol, and takes precedence over
  # `target.dte`
  dte = 60
  # Optional: specify the maximum contract DTE for this symbol (i.e., don't open
  # or roll positions past this DTE). This will override the `target.max_dte`
  # value.
  # max_dte = 45

  # Optional: If we try to roll the position and it fails, just close it (but
  # only if it's profitable).
  close_if_unable_to_roll = true

    # parts = 30
    [symbols.QQQ.puts]
    # Override delta just for QQQ puts
    delta = 0.5
    # Also, optionally specify a strike limit, for either puts or calls.
    # Interpreted as an upper bound for puts, and a lower bound for calls.
    strike_limit = 1000.0 # never write a put with a strike above $1000

    # Optionally, if we only write new contracts when the underlying is green or
    # red (`write_when.*.green=true` && `write_when.*.red=false` or vice versa),
    # specify a minimum threshold as an absolute value daily percentage change
    # (in this example, use 1% for puts only, but could also be specified as
    # `symbols.QQQ.write_threshold`). This can also be specified under
    # `constants.write_threshold` and `constants.puts/calls.write_threshold` to
    # apply to all symbols, either for both puts or calls, or individually.
    #
    # In this example, we'd only write puts on QQQ when the daily change is -1% or
    # greater (provided that we also set `write_when.puts.red=true`).
    write_threshold = 0.01 # 1%, absolute value

    # Alternatively, you can express write threshold value in terms of sigma with
    # `write_threshold_sigma`. A value of 2.0 means that we'll write new contracts
    # when the daily change is twice the standard deviation of the log returns.
    #
    # If write_threshold_sigma is specified, it supersedes (overrides) the value of
    # write_threshold.
    #
    # The daily standard deviation of log returns is calculated based on the past
    # 30 days of data by default, but you can adjust this by altering the value of
    # `constants.daily_stddev_window`.
    #
    # Note that we'll need to retrieve this historical data for every execution to
    # calculate this value, so you should be sure there's 1) sufficient history
    # available and 2) the data can be retrieved sufficiently fast.
    #
    # write_threshold_sigma = 1.0 # 1x the standard devation of log returns for the daily stddev window

      # the values for `write_when.*.green` and `write_when.*.red` can also be set per-symbol
      [symbols.QQQ.puts.write_when]
      green = false
      red   = true

    [symbols.QQQ.calls]
    strike_limit = 100.0 # never write a call with a strike below $100

    maintain_high_water_mark = true # maintain the high water mark when rolling calls

    # These values can (optionally) be set on a per-symbol basis, in addition to
    # `write_when.calls.cap_factor` and `write_when.calls.cap_target_floor.
    cap_factor       = 1.0
    cap_target_floor = 0.0

    # Optionally, only write calls when the underlying has an excess of shares
    # when set to `true`. This overrides the `write_when.calls.excess_only`
    # value.
    excess_only = false

  [symbols.TLT]
  weight = 0.2
  # parts = 20
  # Override delta for this particular symbol, for both puts and calls.
  delta = 0.4

  [symbols.ABNB]
  # For symbols that require an exchange, which is typically any company stock,
  # you must specify the primary exchange.
  primary_exchange = "NASDAQ"
  weight           = 0.05

  # parts = 5
  # Sometimes you may need to wrap the symbol in quotes.
  [symbols."BRK B"]
  # For symbols that require an exchange, which is typically any company stock,
  # you must specify the primary exchange.
  primary_exchange = "NYSE"
  weight           = 0.05

# parts = 5
[ib_async]
logfile = '/etc/thetagang/ib_async.log'

# Typically the amount of time needed when waiting on data from the IBKR API.
# Sometimes it can take a while to retrieve data, and it's lazy-loaded by the
# API, so getting this number right is largely a matter of guesswork.
# Increasing the number would reduce the throughput since we have to wait maximum X seconds
# to make sure all the required data fields are ready.
# You can speed up the thetagang by reducing this number, but you may miss the best opportunities.
# The rule of thumb to choose the best value is assuming the maximum run time of thetagang
# will be around 6 (call,puts,roll calls, roll puts, ...) * api_response_wait_time * number_of_symbols you have in the configuration.
api_response_wait_time = 60

[ibc]
# IBC configuration parameters. See
# https://ib-insync.readthedocs.io/api.html#ibc for details.
gateway     = true
ibcPath     = '/opt/ibc'
tradingMode = 'paper'
# Set this to true if you want to raise an exception on request errors. Under
# normal operation this should be false because we often try to make "invalid"
# requests when scanning option chains for example.
RaiseRequestErrors = false
password           = 'demo'
userid             = 'demo'
# Change this to point to your config.ini for IBC
ibcIni = '/etc/thetagang/config.ini'
# Change or unset this to use something other than the Docker bundled OpenJDK.
javaPath = '/opt/java/openjdk/bin'

# twsPath         = ''
# twsSettingsPath = ''
# fixuserid       = ''
# fixpassword     = ''
[watchdog]
# Watchdog configuration params. See
# https://ib-insync.readthedocs.io/api.html#watchdog for details.
appStartupTime = 30
appTimeout     = 20
clientId       = 1
connectTimeout = 2
host           = '127.0.0.1'
port           = 7497
probeTimeout   = 4
readonly       = false
retryDelay     = 2

  [watchdog.probeContract]
  currency = 'USD'
  exchange = 'SMART'
  secType  = 'STK'
  symbol   = 'SPY'

# Optional VIX call hedging, based on the methodology described by the Cboe VIX
# Tail Hedge Index, described here:
# https://www.cboe.com/us/indices/dashboard/vxth/
[vix_call_hedge]
enabled = false

# Target delta for calls that are purchased
delta = 0.30

# Target DTE for new positions
target_dte = 30

# Optionally specify a maximum DTE for VIX positions, which takes precedence
# over `target.max_dte`.
# max_dte = 180

# If the current spot VIX exceeds this value, long VIX call positions will be
# closed. Comment out to disable.
close_hedges_when_vix_exceeds = 50.0

# Don't count any VIX positions where the DTE is <= this value. Increase this
# value to create a call ladder. For example, if you set this to 5, thetagang
# will ignore current VIX positions starting 5 days before expiry, and
# potentially adding more. This allows you to create a simple call ladder.
ignore_dte = 0

  # The allocations are specified as an ordered list of allocation weights
  # according to an upper/lower bound on VIXMO (the 30 day VIX). Default values
  # are the same as those described in the VXTH methodology. These are evaluated
  # in order, and the weight from the first allocation that matches will be
  # applied. The lower bound is inclusive, and the upper bound is exclusive
  # (.i.e., the code checks that lower <= VIXMO < upper). The upper/lower bounds
  # are only checked if they're present.
  #
  # The allocation weights are multiplied by the account's net liquidation value,
  # and that amount is allocated to purchasing VIX calls.
  [[vix_call_hedge.allocation]]
  upper_bound = 15.0
  weight      = 0.00

  [[vix_call_hedge.allocation]]
  lower_bound = 15.0
  upper_bound = 30.0
  weight      = 0.01

  [[vix_call_hedge.allocation]]
  lower_bound = 30.0
  upper_bound = 50.0
  weight      = 0.005

  [[vix_call_hedge.allocation]]
  lower_bound = 50.0
  weight      = 0.00

[cash_management]
# Cash management gives us a way to earn a little extra yield from excess cash
# sitting in your account. When the cash balance exceeds a threshold, we buy
# the cash fund, and when the cash balance drops below the cash threshold we
# sell the cash fund to get back to the target cash balance.
#
# Enables cash management
enabled = false

# The fund to purchase with your cash. Example of cash funds are SGOV or SHV,
# which are short-term treasury ETFs with reasonable fees. Be sure to check the
# expense ratio before jumping in on ETFs that appear juicier.
cash_fund = "SGOV"

# You don't usually need to specify the primary exchange for ETFs, but if you
# do, you can do so with this:
# primary_exchange = "NYSE"

# The cash balance to target. This is used as a lower bound, so with a value of
# 0, we try not to let the cash balance go below zero. Simple enough.
target_cash_balance = 0

# We don't want to transact too frequently because of commissions, so the buy
# threshold is the amount above target_cash_balance we need to reach before
# placing a buy order.
buy_threshold = 10000

# The sell threshold is the amount below target_cash_balance where we'll place
# a sell order to shore up cash.
sell_threshold = 10000

  [cash_management.orders]
  # The exchange to route orders to. Can be overridden if desired. This is also
  # used for fetching tickers/prices.
  exchange = "SMART"

    [cash_management.orders.algo]
    # By default, use a VWAP order for cash trades. You can comment out the whole
    # `cash_management.orders` section to use the same value as `orders` and
    # `orders.algo`.
    params = [
      # Optionally, uncomment the following line to be opportunistic by avoiding
      # taking liquidity. This gives us somewhat better pricing and lower
      # commissions, at the expense of possibly not getting the order filled in
      # the day.
      # ["noTakeLiq", "1"],
    ]
    strategy = "Vwap"

[exchange_hours]
# ThetaGang can check whether the market is open before running. This is useful
# to avoid placing orders when the market is closed. We can also (for example)
# avoid running too early (i.e., immediately after market open) to avoid the
# initial volatility, or too late (i.e., immediately before market close) to
# avoid the closing volatility.

# The exchange to check for market hours. XNYS is the NYSE, but you can use any
# of the ISO exchange codes from
# https://github.com/gerrymanoim/exchange_calendars?tab=readme-ov-file#Calendars.
#
# For example, you'd use "XTKS" for the Tokyo Stock Exchange, "XLON" for the
# London Stock Exchange, or "XHKG" for the Hong Kong Stock Exchange.
exchange = "XNYS"

# If the market is closed, we can either exit immediately, wait until it's
# time to begin, or continue as normal. Specify "exit" to exit, "wait" to wait,
# and "continue" to continue.
#
# Setting this to "continue" is equivalent to disabling this feature.
action_when_closed = "exit"

# The maximum amount of time, in seconds, to wait for the market to open if you
# set `exchange_hours.action_when_closed = "wait"`. For example, a value of 3600
# means ThetaGang will only wait for the market to open if it's within 1 hour of
# the current time.
max_wait_until_open = 3600

# The amount of time, in seconds, after the market open, to wait before running
# ThetaGang. For example, if this is set to 1800, ThetaGang will consider the
# market closed until 30 minutes after the open.
delay_after_open = 1800

# The amount of time, in seconds, before the market close, to stop running
# ThetaGang. For example, if set to 1800, ThetaGang will consider the market
# closed 30 minutes prior to the actual close.
delay_before_close = 1800