Current ECN rebate rates- Compiling the full list as a useful reference for everyone:

Discussion in 'Order Execution' started by Rearden Metal, Jan 21, 2011.

  1. Let's use this thread to assemble a complete list of all current ECN rebate rates. I'll start building the list. Please correct me if any of the info I post here is outdated.

    This issue is now more confusing than ever, as last year some ECN's completely reversed the old standard practice of paying to add liquidity and charging to remove it. (And I find this completely baffling, btw. How the fuck does paying to remove and charging to add make any sense???)
    ___________________________________________________

    Bats: Very confusing, as 'BATS' can be two entirely different things, as far as I understand.

    BATS BZX: Add +.0027 Remove: -.0028 Route away: -.002 or -.0028

    BATS BYX Add: Free Remove: +.0003 Route away: -.0020 or -.0028
    ______________________

    ARCA: Add Naz or ETF: +.002 Remove Naz or ETF: -.003
    Add NYSE:Free Remove NYSE: Free
    ______________

    EDGX: Add: -.0029 Remove: +.0029 (WTF?! They <i>want</i> their book to be empty & illiquid? Are they on crack, or can someone explain why they'd want to do this?)
    __________________

    EDGA: Add: +.0002 Remove: -.0002

    However, the EDGX & EDGA policies are far more confusing than I've just stated, since there about 40 exceptions to these rates! See: http://www.directedge.com/Membership/FeeSchedule/ECNFeeSchedule.aspx

    Like I said, more confusing than ever...
     
  2. Good idea. But you'll have to stay on top of it. These are subject to change fairly frequently.
     
  3. I think you're off on everything.

    Off the top of my head:
    First is add, second is remove
    Bats: +.0027/-.0028
    BYX: 0/+.0003
    EDGA:-.00025/+.00015
    NQBX:-.0018/+.0014
    EDGX: +.0026/-.003
    NYSE: +.0015/-.0024
    ARCA:+.0021/-.003
    NSDQ:+.0023/-.003
    NQPX:+.0024/-.0013
    LAVA:+.0024/-.0025
    NSX+.0026/-.003

    Hidden liquidity has different rates (-$1/1k for EDGX, -$3/1k EDGA, +$1.0 to add for NSDQ, arca has normal rebate, NSX gives none)

    I'm pretty sure all of these are correct....
     
  4. joe_blo

    joe_blo

    There are so many ECNs (they're almost all actually full-blown exchanges now) that they've got to compete with each other somehow, so they fiddle with pricing and other gimmicks (for example, PHLX is price-size priority; everyone else is price-time).

    If C is sitting there with a million shares offered on every exchange with a penny spread, and someone wants to buy *now*, given the choice, they'll route to an exchange (like EDGA or BOSX) that rebates for removes. The corresponding incentive for someone to pay to add liquidity is a higher probability of execution.
     
  5. +1...

    To be honest I only use a couple of ECNs so I know what their rates are and I just ignore the rest.

    For example why the hell would you use ARCA or NYSE to get filled when BATS gets the fills around about the same time on the thick stocks.

    Just a note though... watch that you use the "ECN"ONLY or "ECN"ALIQ routes otherwise your going to get routed and stuffed on fees. For example it you use EDGXALIQ then you can't remove liquidity. If you just used normal EDGX you may get filled aggressive on hidden EDGX orders at that level. Same goes for BATSALIQ or BATSONLY
     
  6. 7out

    7out

    Hi guys,

    Can you guys also help me by matching ecns to their level 2 codes.

    I know Archipelago lists as PACX on level 2, anyone know the others.

    Thxs in advance.
     
  7. ny212

    ny212

    I'm so confused. Does it mean that ECN is changing fee schedule so that fees minus rebates is positive and maximized? This question has been raised before, but no one really answered it so I'll ask again. If they make it so complicated, wouldn't that drive trading to other exchanges?

    I'm sure DirectEdge has thought it through before making their fee schedule complicated. Could someone in the know enlighten all of us?

    How is complex fee schedule relate to reg NMS?

    Thanks.

    :confused:
     
  8. trom

    trom

  9. so u actually make money by removing EDGA? interesting...
     
  10. drp7804

    drp7804

    I'd be curious to see how the spread for a given symbol on EDGA compares w/ the same symbol on EDGX. Over the long term, I have to believe that EDGA would have a wider spread due to the rebate incentives to both remove liquidity on EDGA and add liquidity on EDGX.

    The thing that makes this tricky is that the rebates/fees are each unique sub-penny amounts. This means that neither ECN can offer the same execution costs for the same contract at any given moment (ie between the two, there will always be one ECN which offers a slightly better deal).

    Hypothetical examples:

    Example A

    On EDGX, stock XYZ bid = 60.00, ask = 60.02
    On EDGA, stock XYZ bid = 60.00, ask = 60.02

    In this case, if you wanted to remove liquidity, the better option would be EDGA. This assumes EDGA offers the same or better likelihood of getting filled (you'd think it'd be better given the rebate to remove).


    Example B

    On EDGX, stock XYZ bid = 60.00, ask = 60.01
    On EDGA, stock XYZ bid = 60.00, ask = 60.02

    In this case, if you wanted to buy at market, EDGX would cost you the spread ($0.01) plus fee ($0.003) = $0.013. EDGA would cost you $0.02 - $0.0015 = $0.0185. So EDGX is the better option.

    Further, if you wanted to do a limit sell at the ask, given equivalent spots in line on both ECNs (probably a big assumption), you'd be better off putting your limit order at $60.02 on EDGA because of the net cheaper entry.

    The catch with each of these examples is that there's some reliance on (necessity for) less-than-perfect trading to keep the machine turning. For instance, in the first example, whoever is offering $60.02 on EDGA could get a better deal on EDGX. In the second example, offering at the ask on EDGA relies on less-than-perfect traders (liq-remove buyers) to not recognize the better deal that EDGX offers them.

    This is all my own hypothesizing, though, as I don't currently have access to that market data... someone with access probably knows the truth.
     
    #10     Feb 17, 2011