I haven't read through the proposal in detail, but since the proposal phase is ending soon, note that fees are DAO parameters. If you have specific numbers in mind, which it looks like you do, you should make DAO parameter change proposals with those targets.

With a generic proposal, even if it's approved by voters in this cycle, you'll still need parameter change proposals to be approved, which won't be possible until the next cycle ends.

@m52go Yes, I would propose the 4 parameters to be changed (BTC Maker Fee, BTC Taker Fee, BSQ Maker Fee, BSQ Taker Fee) but thought I would let people respond.

My background is in stock options market making at the Chicago Board Options Exchange where I worked for 10 years (2000-2010) before retiring due to a combination of being made obsolete by algorithmic trading and stacking enough sats to be free of the modern variety of (corporate) slavery.

Wow, out of all the software engineer cypherpunk geeks here working on this decentralized trading platform, you might be the only actual "trader" around 😉

I don't know the financial incentives at play well enough to conclude for myself if raising trading fees affected the BSQ market over the past few months, or perhaps it was a combination of other factors such as the Bitcoin market in general, or the issues caused by 0.01 BTC trade limits. And the other part of me feels like we are the FED voting to adjust interest rates 😂

But I'm in this for the long run, and I think lowering fees, at least for the short term could be helpful for growth and launching Bisq in new markets. We should be cheaper than centralized exchanges if possible.

I think you should make the 4 parameter changing proposals in the DAO for this cycle and let them be voted on. If it doesn't pass we can always discuss it more and vote again next cycle.

@ifarnung is not alone, I was a market maker for more than 10 years on european stocks. I don´t know if the volume decrease is due to the fee increase, but I share and agree with 100% of his concerns.

However, I have to say that market makers are centralized points for the BTC-fiat pairs, in the sense that banks can easily track most Bisq users by tracking just a few market makers.

Great point @mpolavieja about the market makers being centralized points, and in the case of a banking rail, they will exchange possibly sensitive personal information.

And I'll readily admit that I don't know if that volume decrease is due to fees either, though I do look at a high volume trade like XMR as a canary in the coalmine.

Comparing the Bisq fees to other exchanges they don't seem to be that out of line. The current fees on Bisq, per BTC, are:

BTC maker fee: 0.0020 
BTC taker fee: 0.0060
BSQ maker fee: 13.00
BSQ taker fee: 38.00

With the current BSQ price that's a maker fee around 0.0013 and taker fee at 0.0038, which is higher than what we have been aiming for. Lowering makes sense even with the previous thinking.

Perhaps it's worth lowering more to see if that would drive more trading. In that case I would suggest only lowering the fees in BSQ since that's what any market maker would be using. I suspect that the fees don't matter as much to casual traders compared to the friction of using Bisq at all. My suggested fees would be:

BTC maker fee: 0.0020 
BTC taker fee: 0.0060
BSQ maker fee: 7.00
BSQ taker fee: 21.00

Note that the proposal in the OP seem to have a decimal off for the BTC fees.

@sqrrm I was using a % for the BTC fees but I edited now to the amount of BTC to make it more clear and to match your table.

I agree that to the smaller dollar amount traders, it's likely not worth the effort and expense to acquire around 20-50 BSQ, so maybe a heavy BSQ trading fee discount is the best way to segregate the different trading groups. Both Makers and Takers in a 2.0 BTC trade would be highly incentivized to get some BSQ for the fees.

Trading fees do matter. So we shouldn't be making monthly radical changes about it unless it's really necessary. I didn't like raising fees 100% on the last proposal (although I expected a higher decrease in trading volume) and I wouldn't like to see we cut them by half now.

Traders need to understand that BSQ is very important for Bisq. I'm very positive about making the use of BTC for trading fees very expensive, but do we know % of trades with trading fees paid in BSQ vs BTC? Is it true that small trades are not bothering to use BSQ?

@MwithM Last time we changed the BSQ fees to approximately 0.1% fee for makers and 0.3% for takers. Since BSQ/BTC rate is not fixed we will have to change this fee monthly to keep it at those rates. We should at least lower the BSQ fees to be in line with our target.

feels like we are the FED voting to adjust interest rates

It is more like the contributors decide how much they want to be investors and how much they want to be able to sell of their earned BSQ.

E.g. if fees are too low not much demand from traders is created so contributors will have it harder to sell BSQ and are forced to keep their BSQ as investors. So it is a dangerous trade-off to go too low with the fees. Too high fee -> we lose traders, too low fees -> we lose contributors.

The volume rise over the past 6 months correlates with the global BTC (and XMR) market IMO. It was very bullish until July/August, now it seems we entering a bearish market again (hope not). Bisq had always lower volume when market went bearish and rise in volume when it was bullish. Most Bisq users are more on the long side and holders.

Lets look at current fee costs for a maker:
A 2 BTC (about 16389 USD) trade costs atm 31 USD if BTC is used for fee and 22 USD if BSQ is used. That is 0.189% of trade volume if BTC is used and 0.134% if BSQ is used.
So yes we are a bit over our target (0.1% for BSQ for maker).

I suggest to reduce to:

BSQ maker fee: 10.00
BSQ taker fee: 30.00

We should be cheaper than centralized exchanges if possible.

That is not likely to be feasible. Centralized exchanges are racing to 0 fee (some have already) and doing their business with other revenue streams (altcoin listing costs, and some or many probably frontrunning, user data business,... - and in future they convert to banks and do benefit from that privileged position).

Thanks for the feedback, and it seems there is some consensus around lowering the fees closer to the .1% and .3% target. Based on the current prices, that would be 8-9 BSQ for maker and around 26 BSQ for taker. So I will submit the parameter changes and the DAO stakeholders can vote in the next few days. Thanks for taking the time to respond and I now have a clearer picture of the BISQ vision regarding fees.

that would be 8-9 BSQ for maker and around 26 BSQ for taker.

On which BSQ price is that based on?
Current average is 0.85 USD (as BTC was going down). So I think the 0.1% is about 10 BSQ.

I was using 1 BSQ = 1 $ and the current price of BTC $8,500. .01% = $8.5 .03% = $26.5

I guess I'm now a little confused as I last read that for cycle 6 1.00 USD per 1 BSQ. (bisq-network/compensation#364)

But I see that was at the completion of the last cycle, and if I look today at the DAO tab:

(0.86$)

So what rate should compensation requests be calibrated to?

The price for the cycle was used to have a consensus about the amount requested by all contributors, it was taken at the beginning of the cycle and unfortunately BTC price went down in the past weeks quite a lot, so the BSQ to USD price. For fees I see no reason why we should use such a snapshot price instead using the current price seems reasonable.

Ok, I submitted the parameter changes for just the BSQ Maker and Taker Fee to 10 BSQ and 30 BSQ. Thanks again for all the feedback.

@ifarnung Thank you for a constructive discussion that probably cleared things up a bit for some people. Since we'll have to keep changing the BSQ fee as prices change we'll probably have this discussion again, but with new people. At least then we can point to this discussion.

I also think it would be very interesting to know how much the fee affect different Bisq user types' propensity to trade. Might not be that easy to find out since it's hard to get data on anonymous users :)

Absolutely agree with @ifarnung that we should try to keep fees competitive and market makers happy.

Growing volumes should prioritized over BSQ breakeven for some time. Bisq is still much to small a market to attempt to make it profitable at the expense of slowing down growth.

A little calculation of market maker profits. Of course many premises (volume, frequency, etc) might be wrong... (fees not exact, using 0.5% to be on the safe side).

@flix1 I don't think this parallels reality for the market maker at all. For example, in practice, the premium over spot generally goes down the bigger the amount of BTC. So you may be able to charge 7% over for a .01 (though I pretty rarely see that trade, and only if you are selling BTC, therefore including chargeback risk), it would be tough to sell .5 BTC for 7% over spot, and you would have to wonder at that point if there wasn't some trick on the other end of the trade.

Small point, the capital required would be much more than exactly what you need to post offers, as many ongoing trades can take some days to clear.

But instead of just critiquing a very interesting effort on @flix1 's part (thanks for this!), I'll just post how I think of market making on BISQ. Look at the current best offer in Monero and you see 1.23% over spot.

So these are your costs for going to get the Monero you need to deliver to the client, and this profit assumes a perfectly liquid market on your other exchange. A few ticks of the Monero price, and your $150ish profit can be gone pretty quick. As you see the % the fees are of your profit are quite high and this was why I was arguing to lower the BSQ fees on big 2.0 BTC trades like this (and in general).

But let's see if this proposal was approved, the BSQ fees would be lower by about 30%. :)

@mpolavieja As fee has been adjusted in last voting cycle I guess that can be closed as accepted.

Yeah, I was about to close it and forgot. sorry.