Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

news.ycombinator.com

Post reply on HN
Print view View original

81–90 of 121 posts

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#81

Post by sjbase »

KiranRao0 wrote: post #77

Very cool market. This feels a lot like multi-leg option execution. How do you think market makers are going to react to this? It makes sense for them to provide a bid/ask on individual series, but how do you see them providing liquidity for these more complex orders?

> Very cool market. This feels a lot like multi-leg option execution.

Thanks! And yep, similar but all the way down to the venue/match level, e.g. as opposed to a broker taking on some legging risk to shield the end investor.

> How do you think market makers are going to react to this? Expanding on Kelly's take - the big thing it does for market makers is allow them to manage momentary risk. When a market maker gets filled on an exchange, they are immediately looking to hedge/offload what they took on which involves a sequence of transactions.

Here, the hedge is baked in. So for example, they may enter an order that looks like "buy and/or sell any mix of these 200 securities, if and only if the net change in risk (e.g. change in exposure across several factors) is within some tolerable distance from 0". So that would look like a traditional bid-ask spread across a series of symbols, but with a global exposure constraint. The key outcome being they can quote larger sizes across symbols safely.

NB: the MM doesn't need to know anything about the composition of the complex order on the other side. On top of that, they may be filling one leg, and a natural or other LP filling another etc..

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#82

Post by lpage »

throwmeawaysoon wrote: post #64
matheist wrote: post #63

Can you say more about how you incentivize truthful bidding? I understand about exposure risk and expressive bidding but I'm not sure if that was meant to obviously imply something about truthful bidding which went over my head, or if you meant to not say more about it. I love the part about eventually determining your value-add by comparing to a counterfactual vanilla market -- sounds a bit like Shapley value? If no…

i'm not an economist or a game theorist so i don't remember the details but this paper talks about how certain market designs lead to untruthful bidding https://www.cs.cmu.edu/~sandholm/vickrey.IJEC.pdf but in the context of second price auctions. lpage might be alluding to something having to do with their proxy bidder implementation but the above paper actually discusses how proxy bidders themselves lead to untruth…

VCGs got a real-world test in FB's ad market [1], and the results were mixed. VCG is in a class of theoretically interesting but fragile and overly game-theoretic mechanisms. Our mechanism is boring from a mechanism design standpoint—it's a uniform clearing price periodic auction without any cleaver demand reduction or tricks aimed at incentive compatibility. The complexity of what we allow for with the bidding language makes closed-form/theoretical analysis at best difficult and, in cases, impossible. Instead, we focus on giving traders a direct means to express their valuations and mechanism that minimizes information leakage and post-trade regret (situations where a bidder wishes they'd behaved differently given the auction's outcome).

[1] https://www.researchgate.net/profile/Alexander-Leo-Hansen/pu...

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#83

Post by sjbase »

zitterbewegung wrote: post #71
lpage wrote: post #48

Earlier quoted context omitted.

We're following the project with interest, but for now, we're focused on directing engineering resources to other portions of the stack. FWIW our approach is in the low nanos of precision and accuracy, which puts us within spitting distance of the Nyquist criterion for never aliasing two packets to the same timestamp (thus losing a total ordering) at line rate 10G. That's massive overkill for our purposes (auctions 1…

Well other than GPS being unavailable right ? Another question are your looking into having same day settlement?

Since the GPS signal is just disciplining a local oscillator, it would have to be a sustained outage before drift starts to really matter. But yeah there is a point where it would make a difference.

> Same day settlement

This one is outside our control for the moment - we partner with a 3rd party for clearing and settlement, and would depend on our subscribers also making the switch to same-day.

Once we get into other asset classes, fast settlement is definitely of interest. Some cool stuff we could do with incorporating settlement instructions and/or counterparty risk constraints as part of the expressive bidding language.

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#84

Post by lpage »

throwmeawaysoon wrote: post #79
sjbase wrote: post #75

Earlier quoted context omitted.

> capital injection from YC We raised a series A in 2019 led by Green Visor (who has been excellent btw, and with us from the start) > it seems like you've been building for ~5 years ... i'm guessing it cost quite a bit to develop Yep, you're spot on that it's a complex product. The biggest cost has been making it feel for the user like it's not. What that boils down to is an enormous amount of iterative feedback and…

>We've developed the tech in house i'm not often impressed but that's quite impressive. kudos to you. i currently work on deep learning compilers (as a phd student) but i'm interested in basically all of these things (compilers, combinatorial optimization, auction theory). i know lpage expressed that you're hiring but i'm curious what roles you're hiring for (your careers page is light on details).

We're still a small enough team that we're more focused on talent than roles. As an example of what that means, our stack is polyglot (rust, OCaml, elixir, python), and we don't assume or require that folks have worked in any of those languages before. We invest heavily in learning and teaching.

It sounds like you have a very relevant background, so please email us if you're interested in discussing further!

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#85

Post by jdhn »

lpage wrote: post #78
jdhn wrote: post #73

>Furthermore, bidding in combinatorial auctions can be challenging in both a computational and UX sense. Can you elaborate on how it's challenging in a UX sense? I'm curious to know what the challenges are.

How bidders communicate bids to the auctioneer (the bidding language) is a central concern for any auction. It's pretty straightforward for unit good auctions (I'll pay $5 for A or I'll sell B for $4), but combinatorial auctions involve arbitrary packages of goods; the set of all possible bids is the powerset of the goods being auctioned. Having bidders attach a value to each package is both a computational impossibi…

Thanks for the answer!

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#86

Post by igorkraw »

I sent you an email, but all I can say is please come to Switzerland so I can work with you guys :-) Everything I've seen so far looks quite awesome.

Two questions:

1. Are you implying you are using deep learning heuristics for weighted set packing? Assuming you can't share too much about your IP, did you have a regulatory or business need to deal with worst-case performance guarantees and (how) did you manage this if you did?

2. It sounds like a lot of your stack is OCaml (I'm a fan, 2nd most fanboyed language after Rust and it's a pity it's not more used), is this a deliberate choice or a "grew out of a research project in formal verification where they like ML" consequence?

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#87

Post by renewiltord »

lpage wrote: post #41
renewiltord wrote: post #38

Do you do only equities or also derivatives? This is very interesting. Because you run frequent short auctions, there's no strict long-running orderbook here, right? Are you using FIX for your protocol and where are your servers geographically located?

> Do you do only equities or also derivatives Initially we're US equities only. Stay tuned for other asset classes and geographies. Spot vs derivatives is a core use case that we want to do as soon as we can (only national exchanges can do listed derivatives trades, so it's a big lift). > there's no strict long-running orderbook here, right The default good-till behavior is one auction cycle (100ms Poisson random bac…

Thank you for your answers.

Exciting stuff. And love your docs.

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#88

Post by lpage »

igorkraw wrote: post #86

I sent you an email, but all I can say is please come to Switzerland so I can work with you guys :-) Everything I've seen so far looks quite awesome. Two questions: 1. Are you implying you are using deep learning heuristics for weighted set packing? Assuming you can't share too much about your IP, did you have a regulatory or business need to deal with worst-case performance guarantees and (how) did you manage this i…

Thanks! I'll reply to you there as well.

1. There are two places where deep learning and prior-based approaches can come into play for combinatorial auctions. One is pretty analogous to AlphaZero, but substitute placing a piece on a Go board with accepting a bid, hoping that upon reaching a terminal state the set of bids accepted is feasible and close to optimal. The second is perhaps more in line with what you mentioned—using ML for hyperparameter selection in an algorithm portfolio. When we go live and have production data, our meta optimizer will measure how different approaches are doing and allocate computational resources accordingly in an online fashion. We always use a vanilla unit double auction as a baseline to measure relative performance within an auction cycle, and if the baseline is better, we use it instead.

2. There's a fun and serendipitous story here. I wrote an extremely early prototype as a tiny lisp and evaluator to go with it. We needed a very restricted and functionally pure language that we could control the execution context of, symbolically execute, and do basic formal methods on. The approach worked for a POC, but it was a far cry from real-world adoptable. We proceeded to prototype a DSL with an HM inspired type system and a more pythonic syntax, arriving at a poor man's ML. Better, but a DSL, and something limited/bespoke that would ultimately be annoying for developers. Then we met the guys at Imandra [1], who convinced us that we could have our cake and eat it too using vanilla OCaml/ReasonML and an ultra-high level theorem prover to keep code in an acceptable logic fragment. As an aside, rust is our systems PL and where we do most of the heavy lifting. Evaluating Expressive Bids isn't computationally expensive relative to the optimization problem.

[1]: https://www.imandra.ai/

Re: Launch HN: OneChronos (YC S16) – Combinatorial auctions market for US equities

#89

Post by currymj »

congrats on the launch! We talked briefly a while back, and I've been checking in on your page every so often to see when things would finally get rolling. The world needs more mechanism design.

Recently in school I've been thinking a lot about constant-function market makers. it occurs to me that you can think of a constant-function market maker as being kind of like an expressive bid. That is, putting your assets in a CFMM is saying you're willing to make any trade among a bunch of assets subject to F(net amount of A, net amount of B, net amount of C,..., net amount of $) = k for some F. Regular limit orders are a special case.

Do you have a sense of how your expressive bids overlap with these? Can I cook up some expressive bid that's equivalent to putting assets in a CFMM? What would the restrictions on F be to make things work with your solvers?

Post reply on HN