Maker arbitrage

[cr4zyd3v]

Hey guys.. While I can see this happening on exchanges I was not able to see any topic regarding this methodology..

Usually, From what I could see in many posts, HFT bots search for inefficiencies around market orders and when the bot spots an opportunity, it executes the multiple triangular transactions to get a small %.. However, it is hard to make it in a profitable way due to its competition..

Now, in lots of exchanges out there I can see something like a book war, which make orders being put and removed in order to get the best opportunities..
I am going to test this out by this weekend and see if it works by coding a bot for kucoin.
The logic is something like this:

There are coin A, B and C
lets suppose that B and C are stable currencies just for better explanation
Also, let's not consider fees

So in a random exchange we have the pairs A/B, A/C  and B/C

A/B bid: 0.99  ask: 1.00
A/C bid: 0.98 ask: 1.00

Ok, so if you want to buy 1 of A you have to give 1 of B or C
If you buy and sell it instantly, the best you will get for it is 0.99 of B or 0.98 of C which means a loss trade

What I am going to test is a bot that looks all the time for make positions that worth to be placed

For example: it must be placed close to the book frontier, so it can be fulfilled in a easier manner
The closer of the book frontier, the better the chances to get an order fulfilled but the lower is the % made..

So in the above example the bot would place an bid order around 0.981 and if it gets fulfilled, it could be instantly sold for 0.99 in the other pair

Note that it would not work if it got executed on the market price (taker)..

So here are some facts:
- It is not supposed to make big amounts of money because as you can see, it will depend on low volume trades, as usually it is on book spread area..
- It also depends on the 'lucky' factor because the bot must place the order and get lucky enough to have its order fulfilled before someone puts another order ahead of it (at 0.982 for example..)
- it does not depends of big amounts of money to work.. In theory, it is easy to make 100% of 10$ but not even 0.1% of 100000$ (in a unique order placement)

If you did not know about it, well.. hope you have learned something and let me know what you think!

I am pretty sure that lots of people use it.. Just go in any exchange and take a look on orders with same size on book borders (between ask and bids)

Also, if you have tried or run a bot like this, tips and thoughts are appreciated!

[687_2]

Something to consider is price volatility on different exchanges. If your bot communication is delayed at all with the exchanges, it might get stuck holding a heavy bag when prices begin to move quickly.


A related issue to be aware of is intentional rate limiting. Some exchanges will mess with your API response timing to screw over your trades. Many algo traders have left entirely due to some of the more popular exchanges doing this.

[cr4zyd3v]

Oh yes.. Since this technique works intra exchange, the risk is reduced.. but yess, biggest problem I have seen is rate limitings.. binance for example had a limitation that blocked users in case of placing/removing orders in the book without getting any fulfilled.. still looking for the best scenario 

[pooya87]

Now, in lots of exchanges out there I can see something like a book war, which make orders being put and removed in order to get the best opportunities..

what you see is the representation of people who have no idea how to trade. they buy a crappy bot that has a terrible design which places these stupid orders for them constantly removing and placing again! it is because the users have no plans, not targets,... they just want to compete thinking they can make profit that way!

as for your strategy it relies heavily on a chance that someone fills your placed orders.  that increases the risk of your strategy by a lot. arbitrage itself is a high risk strategy, you may not want to add more to it.