Cheap VPS
press ‘all’ after about 15s
Case in Point
Imagine this hourly graph without the steep drops from paying fees on the green (buy) lines:
As a follow-up to this:
https://hackernoon.com/making-markets-moving-crypto-free-and-open-source-binance-9bcea607e57b
https://github.com/DunnCreativeSS/privateMarketMakerRepo ** ask for access **
TL;DR join us here https://t.me/themarketmakerbot
Elevator Pitch
Market making on Deribit and BitMex failed because it counted on the market to remain more or less stagnant on the 0.25$ step. We’re now looking at automating two market making strategies for smaller volume, higher spread pairs.
Market Maker Trader
We buy just above the highest bid and sell just below the lowest ask. We repeat this process, using a fraction of account balance in base pairs, until we see the net profits as price fluctuations cancel each other out in the approaching infinity sense — while we soak up profits from the spread itself in the ‘relative orders’ strategy.
Conversationally, and not coded yet, is the ‘staggered order’ strategy, where you pick a maximum and minimum price for that pair and then stagger orders up and down the order book to buy and sell along set intervals.
Resources: Would need servers close to different exchanges.
Scalability: More coins; more scale
Pros: I’ve had 0.57 bitcoin volume on my deposit of $18 worth of coins on the ‘relative’ strategy in the last 24 hours, while sustaining about -0.5% growth. There are other exchanges (some with margin) where we can reproduce the bot, like bitfinex/ethfinex who have a market maker rebate paid monthly in their proprietary coin, or liquid who has a market maker rebate on pairs that were previously on qryptos
Add’l pros: on Binance I can effect a 20% or 40% income on fees via my referral link, on HitBTC the affiliate program is on pause but I may eventually be able to effect 75% — although the potential gains from HitBTC affiliate are less as there will eventually be a 0% or rebate on the maker fees
Cons: To prove profitable on Binance or HitBTC, you’d need to eliminate the fees or effect a market maker rebate. On Binance this involves volume as well as holding BNB — while still paying some fees, while on HitBTC that only involves building volume first — effecting 0% maker fees then 0.01% rebate. https://hitbtc.com/fee-tier https://www.binance.com/en/fee/schedule On HitBTC there’s also a market making program https://hitbtc.com/mm.
To Prove for Viability
Consistent over all types of markets, but identify when it does better
In the long run, price volatility helps the ‘staggered’ strategy more than it does the ‘relative’ strategyLess volatility but still having volume, there will be gains from the ‘relative’ strategyThe potential loss is 1. Fees 2. Grabbing a coin that immediately dies, inclusive.If you’re to eliminate fees or effect a fee rebate, this risk goes downIf you automate many pairs you lessen the negative effect of grabbing a coin that diesConversationally, a stop loss can be created to further lessen this risk
Sunk costs of development
None — need to code the ‘staggered’ strategy, which I can do
Scalability given current market liquidity and volume
On exchanges like Binance or HitBTC, with many coin pairs that have significant volume, we can scale indefinitelyThe bot currently checks the average volume per base asset, then the spread of a given pair, and enters only into those markets that meet the minimum and maximum volumes and target spread, as well as a minimum order quantity and maximum order quantity (to avoid sh#tcoins)
In base: market pair: volume in base, that looks like this:
{ BTC: { XDNBTC: 69.3393775486, VETBTC: 42.606415325 }, TUSD: { NXTUSD: 67247.7248096, TNTUSD: 53891.073732 }, MUSD: { QTUMUSD: 2545776.043225 }, NUSD: { XDNUSD: 280614.7629689 }, DUSD: { MAIDUSD: 54019.8292817 }, ETH: { ICXETH: 589.145211656, NXTETH: 266.992269315, REPETH: 616.456689351, ONTETH: 5359.788369161, NTKETH: 197.416308344, KBCETH: 355.311810444, ROXETH: 311.845628547 }, BUSD: { DGBUSD: 2445.14526651 }, GUSD: { BTGUSD: 141765.29622322 }, UUSD: { QNTUUSD: 29.4731369 }, PUSD: { ETPUSD: 68.5015549, ZAPUSD: 40.1461164 }, YUSD: { DAYUSD: 338.8757426, BERRYUSD: 248.9052084 }, QUSD: { STQUSD: 6039.7474451 }, IUSD: { WIKIUSD: 10575.2656811 }, FUSD: { ELFUSD: 14.04120135 }, URS: { ETHEURS: 2486.24206583, LTCEURS: 1485.789705, XMREURS: 2807.0239136 }, ‘0USD’: { POA20USD: 20.1087685 }, EOS: { LSKEOS: 39.06901764 }, RWB: { BTCKRWB: 44.82036 } }
Forward Tests
This strategy was first coded about 48 hours ago on Binance. It lost about 0.5% in one day, and had 0.46 BTC in volume.
The second iteration was on HitBTC, and as of about 10–12 hours ago has 0.12 BTC in volume (across many smaller orders, instead of the test version on Binance risking everything on one pair).
It’s lost about 0.23% in that time:
My calculations, based on losing 0.5% a day on average with entry level fees and based on HitBTC’s 0% maker fee after 1500BTC in 30-day volume, indicate that after a certain amount of time while trading a certain balance we can effect 0% fees, and therein effect profits:
https://docs.google.com/spreadsheets/d/1U343CX4ghe_Au9Q3QwHZHW8A0WIwwprgP23mndKOzAE/edit?usp=sharing
This is not including the chance we can get market maker benefits or the 0.01% rebate after 6000 BTC volume.
Conclusion
With enough volume anything is possible. Come join us! https://t.me/themarketmakerbot