---
title: "Getting Access: Crypto Venues"
book: "Markets III: Commodities, Energy and Crypto"
subject: quant
language: en
chapter: 25
exercises: 8
source: https://one-course.com/books/quant/3/en/chapter/25-getting-access-crypto-venues
---

# Chapter 25 — Getting Access: Crypto Venues

A new trading firm wants to quote on five crypto venues. It opens accounts, connects, and trades. At the end of its first month it finds that it paid, per dollar traded, many times what an established market maker pays for the same trades, because fees fall with volume and its volume is spread thin; that its connections were throttled at a fraction of the rate its strategy needed; and that the rebate it was counting on went to firms that kept their quotes up a few percentage points more of the time. Access to a crypto venue is cheap to obtain and expensive to use well. This chapter completes the access chapters of the series for crypto: the entities and onboarding a venue requires, fee schedules and how to be on the right tier, market-maker programmes and what they demand, connectivity, and the credit and custody arrangements that let a firm trade without leaving its assets on venues.

## 25.1 Entities, onboarding and sub-accounts

**Definition 25.1 (Know-your-customer check).**

A *know-your-customer check* is a venue’s verification of a client’s identity, ownership and control, source of funds and regulatory status before it opens an account, repeated periodically and when facts change.

An institutional account on a crypto venue is opened by a legal entity, and the venue’s onboarding asks what a broker’s would (One Quant Book 1, chapter 29): incorporation documents, the entity’s legal entity identifier (One Quant Book 2, chapter 28), the beneficial owners and directors (Kraken’s business verification asks for formation documents, a share registry of ultimate beneficial owners with their capital and voting rights, an ownership chart and a proof of operating address), the source of the initial funds, the firm’s regulatory licences, and its trading strategy. The venue’s own jurisdiction decides whom it may accept: an offshore venue may exclude residents of countries where it is not licensed, and a licensed venue may require the client to be licensed too ([Chapter 24](https://one-course.com/books/quant/3/en/chapter/24-the-crypto-trading-business#ch-m3-the-crypto-trading-business)). Once open, the master account is divided into [sub-accounts](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-subaccount) ([Chapter 15](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#ch-m3-centralised-exchanges)) by strategy, each with its own [API keys](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-subaccount) and limits.

## 25.2 Fee schedules, VIP tiers and tier matching

Crypto venues charge by the volume tiers of One Quant Book 1, with maker and taker rates in basis points and tiers set on the thirty-day trading volume of the account (sometimes of all its [sub-accounts](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-subaccount) together), and sometimes on holdings of the venue’s [token](https://one-course.com/books/quant/3/en/chapter/14-blockchains-for-traders#def-m3-blockchains-for-traders-key).

**As of September 2026 — Two spot fee schedules in full.**

Binance’s spot schedule runs from the regular tier (under USD 1 million of thirty-day volume, 0.100% maker and taker) through VIP 1 (USD 1 million, 0.090%/0.100%), VIP 2 (5 million, 0.080%/0.100%), VIP 3 (20 million, 0.040%/0.060%), VIP 4 (75 million, 0.040%/0.052%), VIP 5 (150 million, 0.025%/0.031%), VIP 6 (400 million, 0.020%/0.029%), VIP 7 (800 million, 0.019%/0.028%) and VIP 8 (2 billion, 0.016%/0.025%) to VIP 9 (4 billion, 0.011%/0.023%), each tier also listing a BNB holding. Kraken Pro’s spot schedule runs from 0.40%/0.80% below USD 2 500 of thirty-day volume through 17 tiers, reaching a 0.00% maker fee from USD 10 million and 0.00%/0.05% from USD 500 million.

Tiers make a firm’s cost a function of how it allocates its volume. Fees fall with volume on each venue, so spreading a fixed volume over five venues leaves each on a lower tier than concentrating it on two.

![The blended fee (70% of volume as maker, 30% as taker) on the two published schedules of as a function of thirty-day volume. Each step is a tier: below USD 2.5 million Binance’s rates are the lower, above it Kraken’s, whose maker fee falls to zero from USD 10 million. Data: the chapter’s tutorial.](https://one-course.com/images/onecourse/chapters/quant-3/m3-getting-access-crypto-venues/fig-4725a4988ab4.svg)

***Figure 25.1.** The blended fee (70% of volume as maker, 30% as taker) on the two published schedules of [Box 25.1](#dat-m3-getting-access-crypto-venues-tiers) as a function of thirty-day volume. Each step is a tier: below USD 2.5 million Binance’s rates are the lower, above it Kraken’s, whose maker fee falls to zero from USD 10 million. Data: the chapter’s tutorial.*

**Proposition 25.2 (Concentration lowers the fee bill).**

If every venue’s schedule is non-increasing in volume and the firm’s volume and maker share are fixed, moving all the volume of one venue to another that already has a lower rate at its current volume never raises the total bill; the saving is at least the old venue’s volume times the difference in rates.

**Proof.** The moved volume now pays the receiving venue’s rate, at most its old rate there since the receiving venue’s volume rose; the receiving venue’s existing volume pays a rate that did not rise; the old venue’s bill vanishes. ∎

![A month of USD 600 million, 70% as maker, spread evenly over five venues (USD 321 000 of fees) or concentrated on two (USD 143 000). Binance and Kraken on the schedules of ; venues C, D and E on illustrative schedules. Data: the chapter’s tutorial.](https://one-course.com/images/onecourse/chapters/quant-3/m3-getting-access-crypto-venues/fig-09fb6f7f5262.svg)

***Figure 25.2.** A month of USD 600 million, 70% as maker, spread evenly over five venues (USD 321 000 of fees) or concentrated on two (USD 143 000). Binance and Kraken on the schedules of [Box 25.1](#dat-m3-getting-access-crypto-venues-tiers); venues C, D and E on illustrative schedules. Data: the chapter’s tutorial.*

**Definition 25.3 (Tier matching).**

*Tier matching* is a venue’s offer to grant a new client, for a trial period, the fee tier the client holds on a competing venue, on evidence of that tier, so that it need not first trade up the schedule.

[Tier matching](#def-m3-getting-access-crypto-venues-match) is how venues compete for flow that has already concentrated elsewhere. For a firm it changes the arithmetic of the proposition: a third venue that matches its best tier costs little to add, and the saving from concentration becomes a question of whether the matched tier survives the trial.

**As of September 2026 — A tier match and a venue’s liquidity programmes.**

From 15 September to 31 October 2026 Bitget offered VIP traders of other exchanges a thirty-day card for its top tier on proof of at least 50 million USDT of futures volume or 30 million of spot volume over the previous thirty days. Bybit’s published market-maker programme (September 2026) pays spot maker rebates of 0.1 basis point above 25 million USDT of monthly volume and 0.5 basis point from a 0.1% share of all market makers’ maker volume, weights altcoin perpetuals eight times in the share that sets perpetual tiers, requires orders of at least ten times the minimum size, reviews levels monthly and may withdraw every benefit from a market maker who misses them; its public spot schedule runs from 0.10% maker and taker to 0.030% and 0.045% at 100 million of monthly volume. Binance announced on 4 June 2025 an altcoin programme paying 0.5 and 1 basis point to providers reaching 0.5% and 1% of maker volume in selected [tokens](https://one-course.com/books/quant/3/en/chapter/14-blockchains-for-traders#def-m3-blockchains-for-traders-key), next to a spot maker programme paying up to 0.8.

## 25.3 Market-maker programmes and their obligations

Beyond the public schedule, venues sign market makers to programmes: rebates on maker volume, sometimes a fixed fee, in exchange for quoting obligations of the kind One Quant Book 1 describes for designated market makers (chapters 24, 29 and 30), measured by the venue on its own data.

**Definition 25.4 (Uptime requirement).**

An *uptime requirement* is the minimum share of sampled moments, over a measurement period, at which a market maker’s quotes in an instrument must meet the programme’s conditions on both sides: a maximum spread and a minimum size within a band around the mid price.

**Proposition 25.5 (Uptime is decided in the tails).**

If a strategy’s quotes fail the programme’s conditions whenever the market’s volatility exceeds a threshold $v^*$, its uptime is $\Pr(v < v^*)$, and raising the uptime from $1 - p$ to $1 - p'$ requires quoting through the volatility quantile $1 - p'$: the quiet hours contribute nothing to the margin.

**Proof.** The quotes pass at exactly the moments with $v < v^*$; the uptime is the frequency of those moments, a quantile of the volatility distribution. ∎

![Uptime of a quoting strategy against a programme requiring USD 50 000 a side, as a function of the maximum spread allowed. The strategy quotes 8 basis points in normal conditions, widening with volatility, halves its size in volatile moments and pulls its quotes above three times normal volatility. Keeping full size up to 2.25 times normal volatility instead of 2.0 lifts uptime from 87% to 91%, past a 90% requirement (dotted). Simulated minutes. Data: the chapter’s tutorial.](https://one-course.com/images/onecourse/chapters/quant-3/m3-getting-access-crypto-venues/fig-c3beab60d68a.svg)

***Figure 25.3.** Uptime of a quoting strategy against a programme requiring USD 50 000 a side, as a function of the maximum spread allowed. The strategy quotes 8 basis points in normal conditions, widening with volatility, halves its size in volatile moments and pulls its quotes above three times normal volatility. Keeping full size up to 2.25 times normal volatility instead of 2.0 lifts uptime from 87% to 91%, past a 90% requirement (dotted). Simulated minutes. Data: the chapter’s tutorial.*

The requirement bites exactly when quoting is dangerous: a firm that pulls its quotes in the volatile minutes, which is where its adverse selection is worst, loses the rebate for the whole month. The programme pays for risk taken in the tails, and the firm must decide whether the rebate is worth it, instrument by instrument.

## 25.4 Rate limits and dedicated endpoints

**Definition 25.6 (Dedicated endpoint).**

A *dedicated endpoint* is a network entry point to a venue reserved for a class of clients, typically market makers or high-tier accounts, with higher [rate limits](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-ratelimit) or a shorter path to the matching engine than the public endpoints.

The [rate limits](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-ratelimit) of [Chapter 15](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#ch-m3-centralised-exchanges) set a firm’s quoting budget, and the tightest rule decides it. At the limits of [Box 15.1](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#dat-m3-centralised-exchanges-limits), the order rule of 100 per 10 seconds allows ten messages a second, but the daily rule of 200 000 allows only 2.3 a second sustained: a firm quoting five instruments can change each quote about once every two seconds on average, unless it negotiates higher limits. Venues also publish several public endpoints of different characters.

**As of September 2026 — A venue’s public endpoints.**

Binance’s spot API documentation lists base endpoints api.binance.com, api-gcp.binance.com and api1 to api4.binance.com, noting that the last four “should give better performance but have less stability”, a separate endpoint for public market data only, and a request validity window (`recvWindow`) of 5 000 milliseconds by default and 60 000 at most.

## 25.5 Loan lines, off-exchange settlement and custody

**Definition 25.7 (Loan line, collateral mirroring).**

A *loan line* is a credit facility from a venue or its affiliate that lets an institutional client borrow assets or trade on margin beyond its deposits, against collateral and a credit assessment. *Collateral mirroring* is the crediting, on a venue, of a trading balance equal to collateral that the client keeps with a custodian, which is the mechanism of [off-exchange settlement](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-oes) ([Chapter 15](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#ch-m3-centralised-exchanges)).

The access decision is therefore also a credit decision in both directions. The firm extends credit to the venue by depositing; the venue extends credit to the firm through [loan lines](#def-m3-getting-access-crypto-venues-loan) and mirrored balances, and prices it in the limits and haircuts it applies. A firm that trades on five venues with [off-exchange settlement](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-oes) through one custodian holds most of its collateral in one place, nets its exposures across venues, and replaces five venue exposures with one custodian exposure and the unsettled balances in between.

## 25.6 Tutorial: five venues, one budget

**Goal.** Given a month’s volume and maker share across five venues and their fee schedules, compute each venue’s fees, the saving from concentrating volume, the value of a tier match, and the uptime a strategy achieves against a programme’s conditions. **End state:** Figures [25.2](#fig-m3-getting-access-crypto-venues-fees) and [25.3](#fig-m3-getting-access-crypto-venues-uptime) and the numbers of the weekend problem.

1. **Tiers and fees.** The tier for a volume, a month’s bill, and a matched tier. `@dataclass (frozen=True ) class VolTier : min_volume: float # thirty-day volume, USD maker_bp: float taker_bp: float def tier_for (volume: float , tiers: list [VolTier]) -> VolTier: return max ((t for t in tiers if volume >= t.min_volume), key=lambda t: t.min_volume) def monthly_fees (volume: float , maker_share: float , tiers: list [VolTier], matched: VolTier | None = None ) -> float : """Fees in USD for a month; a tier match, if granted, replaces the tier when it is better.""" t = tier_for(volume, tiers) if matched is not None and matched.maker_bp + matched.taker_bp < t.maker_bp + t.taker_bp: t = matched return volume * (maker_share * t.maker_bp + (1 - maker_share) * t.taker_bp) / 1e4` **Listing 25.1.** Volume tiers, a month’s fees and a tier match. code/firm/mmprogram/firm_mmprogram.py
2. **Uptime.** Sampled snapshots of the firm’s own quotes checked against spread and depth. `def compliant (s: Snapshot, max_spread_bp: float , min_depth_usd: float ) -> bool : if s.bid is None or s.ask is None : return False return 1e4 * (s.ask - s.bid) / s.mid <= max_spread_bp and min (s.bid_depth, s.ask_depth) >= min_depth_usd def uptime (snaps: list [Snapshot], max_spread_bp: float , min_depth_usd: float ) -> float : return sum (compliant(s, max_spread_bp, min_depth_usd) for s in snaps) / len (snaps) def rebate (maker_volume: float , rebate_bp: float , measured_uptime: float , required: float ) -> float : """Programme rebate in USD: paid on all maker volume if the uptime requirement is met, else nothing.""" return maker_volume * rebate_bp / 1e4 if measured_uptime >= required else 0.0` **Listing 25.2.** Compliance of a snapshot, uptime and the programme rebate. code/firm/mmprogram/firm_mmprogram.py
3. **Run** `spread_vs_concentrate()` , `tier_match_value()` , `programme_value` and `fig_cryptoaccess.py` .

**What to change next.** Find the allocation over the five venues that minimises fees subject to at least 10% on each (for resilience); add a programme’s fixed fee; recompute uptime when the [rate limit](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-ratelimit), not volatility, is what makes quotes stale.

## 25.7 Build: the market-maker programme monitor

**Purpose.** The miniature firm must know, every day, which tier it is on at each venue and what its allocation costs, whether it is meeting each programme’s obligations, and how many quote changes its limits allow.

**Interface.** `VolTier(min_volume, maker_bp, taker_bp)`; `tier_for(volume, tiers)`; `monthly_fees(volume, maker_share, tiers, matched)`; `Snapshot(mid, bid, ask, bid_depth, ask_depth)`; `compliant`; `uptime`; `rebate`; `requote_budget(governor, instruments)` from `firm.ratelimit`.

**Rules.** Schedules versioned by date; volume counted as the venue counts it; snapshots of the firm’s own quotes, sampled as the venue samples; a programme’s rebate booked only when its measurement confirms it.

**Acceptance tests.** `code/firm/mmprogram/tests/`: tiers and a month’s bill; a tier match that helps only when better; compliance, uptime and the all-or-nothing rebate; the requote budget set by the tightest order rule.

**Stretch.** An optimiser for the allocation of volume across venues; intraday uptime alerts; the programme’s measurement reproduced from the venue’s market data.

Sources and further reading

- Binance, spot fee schedule and spot API documentation (general information, limits); Kraken, fee schedule. Accessed September 2026.
- Copper, ClearLoop product page (see [Chapter 15](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#ch-m3-centralised-exchanges) ).
- Kraken, support article “How to get verified for a Business account” (updated 24 August 2026); Bitget, press release of 16 September 2026; Binance, press release of 4 June 2025 (PR Newswire); Bybit, help centre, “Introduction to the Market Maker Incentive Program” and “Bybit Trading Fee Structure”, September 2026.

## 25.8 Exercises

**Exercise 25.1 ★.**

On Binance’s schedule, what does a month of USD 100 million, 70% as maker, cost?

**Solution of Exercise 25.1.**

USD 100 million is on VIP 4 (0.040%/0.052%): $100\text{ million} \times (0.7 \times 4 + 0.3 \times 5.2)$ basis points, USD 43 600.

**Exercise 25.2 ★.**

What does a venue’s [know-your-customer check](#def-m3-getting-access-crypto-venues-kyc) ask of a trading firm, and why does the venue care?

**Solution of Exercise 25.2.**

Identity and incorporation, beneficial owners and directors, source of funds, licences and regulatory status, the trading strategy. The venue bears the legal risk of serving a sanctioned or unlicensed client or laundering funds, and its own licences depend on these checks.

**Exercise 25.3 ★.**

Why does the daily order rule, not the ten-second one, set a quoting firm’s sustained budget at the limits of [Box 15.1](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#dat-m3-centralised-exchanges-limits)?

**Solution of Exercise 25.3.**

Ten a second for ten seconds is 100, but 200 000 a day is only 2.3 a second on average: a strategy that quotes all day runs into the daily rule long before it saturates the short one.

**Exercise 25.4 ★★.**

Venue C matches the firm’s Binance tier on its USD 120 million there. What is the match worth a month?

**Solution of Exercise 25.4.**

USD 11 040 a month: from 3.6 basis points on venue C’s own schedule to 2.68 on the matched Binance tier.

**Exercise 25.5 ★★.**

A programme pays 0.5 basis points on maker volume at 90% uptime. On USD 210 million of maker volume, what is it worth, and what is the cost of missing the requirement by one point?

**Solution of Exercise 25.5.**

USD 10 500 a month; missing by one point loses all of it, since the rebate is paid only if the requirement is met.

**Exercise 25.6 ★★.**

Why might a firm keep 10% of its volume on each of five venues even though concentrating it is cheaper?

**Solution of Exercise 25.6.**

Resilience and information: a venue can fail, suspend withdrawals or go down in a crisis, and prices and liquidity on each venue inform quoting on the others; small allocations also keep the accounts, connections and relationships alive.

**Exercise 25.7 ★★★.**

*Coding.* With `programme_value`, find the uptime at maximum spreads of 10 and 20 basis points when depth is halved above 2.5 times normal volatility.

**Solution of Exercise 25.7.**

About 64.0% at 10 basis points and 93.6% at 20: at 10, the spread limit binds long before the depth rule does.

**Exercise 25.8 ★★★.**

*Find the flaw.* “We will reach the VIP tier faster by trading with ourselves across two [sub-accounts](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-subaccount).”

**Solution of Exercise 25.8.**

Trading between one’s own [sub-accounts](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-subaccount) is [wash trading](https://one-course.com/books/quant/3/en/chapter/16-spot-markets#def-m3-spot-markets-wash); venues exclude it from tiers where they can detect it, it breaches their terms, and it may be illegal manipulation. It also pays taker fees on half of it.

## 25.9 Problem: Five Venues, One Budget

**Problem 25.1.**

Weekend problem — allocating a month of volume

A firm trades USD 600 million a month, 70% as maker. It can spread the volume evenly over Binance, Kraken and three other venues (C, D and E, on the tutorial’s illustrative schedules), or concentrate it on Binance and Kraken. One venue offers a programme paying 0.5 basis points on maker volume for 90% uptime at a maximum spread of 20 basis points and USD 50 000 a side.

**Part I — The bill.**

1. What does each venue cost when the volume is spread evenly, and in total?
2. What do the two venues cost when the volume is concentrated?
3. What is the monthly saving?
4. Which venues’ tiers change, and why is Kraken so much cheaper for a maker?
5. What would a tier match at venue C change?

**Part II — The programme.**

6. On USD 300 million at the programme’s venue, what is the rebate worth?
7. What uptime does the firm’s strategy achieve, and does it qualify?
8. What change to the strategy lifts it past 90%?
9. What does that change cost in risk?
10. How does the requirement interact with the venue’s [rate limits](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-ratelimit) ?

**Part III — Beyond fees.**

11. What exposure to the venues does each allocation create?
12. How would [off-exchange settlement](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-oes) change the comparison?
13. What does the firm give up by leaving three venues?
14. How should it measure where its trades are best executed?
15. How often should it revisit the allocation?

**Part IV — Judgement.**

16. Is concentrating volume good for the market?
17. Why do venues pay rebates rather than cut fees for everyone?
18. What is the right number of venues for a new firm?
19. State the *named result* : the monthly saving from concentrating on two venues, and the uptime the firm must keep, and currently keeps, for the rebate.
20. In one sentence: what does a crypto venue sell a market maker?

**Solution of Problem 25.1.**

**1.** USD 52 320 (Binance), 28 800 (Kraken), 43 200 (C), 86 400 (D) and 110 400 (E): USD 321 120. **2.** USD 80 400 and 63 000: USD 143 400. **3.** USD 177 720 a month. **4.** Binance moves from VIP 4 to VIP 5 and Kraken from the USD 100 million to the USD 250 million tier; Kraken’s maker fee is zero from USD 10 million, so 70% of its volume pays nothing. **5.** It would cut C’s cost by USD 11 040, making a three-venue allocation cheaper than spreading and closer to concentration. **6.** $300\text{ million} \times 70\% \times
0.5$ basis points: USD 10 500 a month. **7.** 87.3%: it does not qualify. **8.** Keeping full size up to 2.25 times normal volatility instead of 2.0: 91.0%. **9.** Full-size quotes in more of the volatile minutes, when adverse selection is highest; the rebate must exceed the extra expected losses. **10.** Keeping quotes compliant in fast markets needs updates, which the order-count rules ration; the firm must budget its messages for the moments the programme measures. **11.** Concentration puts USD 300 million a month through each of two venues, and the collateral and balances there; spreading halves each exposure but adds three. **12.** Collateral held with a custodian and mirrored on the venues reduces what each venue failure can take, making concentration of trading less of a concentration of credit. **13.** Liquidity and prices on those venues, a fallback if one of the two fails, and the option to trade there later. **14.** By the realised cost of its own trades on each venue (spread paid, slippage, markouts), not by fees alone. **15.** Monthly, and whenever a schedule, a programme or its volume changes. **16.** It lowers the firm’s costs and deepens two books, but it thins the others and concentrates the market’s liquidity and risk. **17.** Rebates buy the quoting behaviour the venue wants from the firms that can provide it; cutting fees for everyone buys nothing specific. **18.** Few: two or three where it can reach good tiers, plus one kept small for resilience. **19.** *Named result:* USD 177 720 a month saved by concentrating on two venues; the rebate requires 90% uptime, which the strategy reaches (91.0%) only if it keeps full size up to 2.25 times normal volatility, against 87.3% as run. **20.** Access to its customers’ order flow, on terms that improve with the volume and the risk the market maker brings.

## 25.10 Interview questions

**Interview question 25.1 ★ trader.**

How do volume tiers change where you should trade?

**Solution of Interview question 25.1.**

Fees fall with volume on each venue, so concentrating volume lowers the average fee; the cheapest allocation puts most volume where the schedule rewards it, subject to resilience, liquidity and credit limits per venue.

*What the interviewer is looking for: marginal tier effects and non-fee constraints.*

**Interview question 25.2 ★ risk.**

What would you check before depositing USD 20 million on a new venue?

**Solution of Interview question 25.2.**

Its licences and jurisdiction, ownership, financial statements or [proof of reserves](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-por) and liabilities, custody and segregation terms, withdrawal track record, insurance, its treatment of client assets in insolvency, and the firm’s credit limit and sweep policy for it.

*What the interviewer is looking for: credit analysis of the venue.*

**Interview question 25.3 ★★ developer.**

How do you design a quoting system to meet an [uptime requirement](#def-m3-getting-access-crypto-venues-uptime) without exceeding [rate limits](https://one-course.com/books/quant/3/en/chapter/15-centralised-exchanges#def-m3-centralised-exchanges-ratelimit)?

**Solution of Interview question 25.3.**

Measure uptime continuously as the venue does; prioritise quote updates on the instruments at risk of failing; coalesce updates; keep a message budget for fast markets; widen within the allowed spread rather than cancel; alert before the monthly measure drops below the threshold.

*What the interviewer is looking for: measurement, prioritisation and a message budget.*

**Interview question 25.4 ★★ trader.**

Is a market-maker rebate worth taking if it requires quoting through volatile periods?

**Solution of Interview question 25.4.**

Only if the rebate exceeds the expected adverse-selection losses of quoting through those periods, measured from the firm’s own markouts in volatile minutes; the answer differs by instrument.

*What the interviewer is looking for: rebate against measured tail losses.*

**Interview question 25.5 ★★ researcher.**

How would you measure whether a venue’s volume is real before joining its programme?

**Solution of Interview question 25.5.**

Compare its volume with its [quoted depth](https://one-course.com/books/quant/3/en/chapter/16-spot-markets#def-m3-spot-markets-depth) and web or on-chain activity, test its trade sizes for rounding and Benford’s law ([Chapter 16](https://one-course.com/books/quant/3/en/chapter/16-spot-markets#ch-m3-spot-markets)), check its prices’ co-movement with the main venues, and trade small amounts to see the real depth.

*What the interviewer is looking for: independent evidence of real activity.*

**Interview question 25.6 ★★★ developer, trader.**

Build the daily report a head of trading needs on the firm’s five venue relationships.

**Solution of Interview question 25.6.**

For each venue: volume and tier, fees and rebates earned, uptime against each programme, message usage against limits, balances and collateral against credit limits, withdrawal and connectivity incidents, and execution quality; with alerts for threshold breaches and a monthly allocation recommendation.

*What the interviewer is looking for: costs, obligations, limits and credit in one view.*