Experiment of Trading Volume at Different Spreads Using a Gateway

Can you think of a different strategy on an open exchange where you change price without cancelling the order?

How about always placing order in two equal halves, and when moving, moving one half first then the second. So there is always some liquidity at one of the two walls?

For the record, volume of nbt/btc on polo in the last 24hr is 2369 NBT.
@willy, what are these numbers on https://alix.coinerella.com/volume/gui_index.php ? say , Alix 7 means ?

Note: The columns value shows the number of stored data arrays for the pair in said time frame and is currently used for debugging purposes only. We check whether our gathered data doesn’t contain more than one data array per day. In this example the total volume for 7 days was 74084.68313822.This value is divided by the number in the columns value, which should be identical with the frame value, to get the ALix result.

https://coinerella.com/index.php/ALix

I totally don’t get what you say. Can you given an example?

ALix 7 = sum(liquidity of last seven days)/seven

And the number in the table is sum(liquidity of last seven days) ?

wait don’t you mean sum(volume of last seven days) ?

ALix 7 = all volume values for the last seven days e.g… 10+20+30+40+50+60+70 dived by 7

It’s a moving average of the past x days

So currently alix7 is 16987.0262 for polo. It means the average volume on polo in the last 7 days is 16987.0262 nubits, right?

1 Like

There are zero trades since my last report.

1 Like

How is now? the first 4hr after the spread was set to 1%, right?

Still no trades on Poloniex. Even with the 1% spread, there are multiple walls in front of me most of the time. Probably MoD’s gateway, the ALP and NuLagoon. Together these walls seems to leave no space for trades at higher spreads on the sell side.

you are right. We have preliminary data – your trading data and price data – to show that sell walls with small spread don’t help to bring volume.

Please keep the spread constant. We want to take more data point and observe dynamic effects when btc moves.

If the sell prices are undercut by 2% or more, this explains why my NuBot at 2% buyside offset traded into existing orders when it placed its orders.

To me it sounds like we need to increase buyside offset to 4% for NuOwned operations at a tight buyside offset and 6 to 7% for those with an increased offset.
At the current order situation ar Poloniex, @Cybnate’s PyBot is pretty much useless except for pumping BTC into the market.

Small spreads with only sell walls are not likiely working as long as there are no buy order. The Poloniex market depth graph clearly indicates that when you sell you would have a hard time to buy back BTC. Not attractive. I think the experiment requires reducing buy side spreads slowly, say 0.5% daily to see where the threshold is for market uptake.

At the current Poloniex situation, you’d need like 3% buyside offset and -2% sellside offset to have competitive orders at 1% spread.

We never had a problem with that before. Probably because we set the market. It is now set by others which shows frictions in correct price determination, not unusual.

To me it sounds like we should decrease offset to find out where the actual front line is. Our customers needs our support.

I agree. We are in unsailed waters and we should experiment with small amount.

1 Like

These posts have been separated from the current liquidity thread as asked by @mhps.

@Cybnate, any news with the 1% gateway?