--- title: makershire hatherway draft tags: practice MIPs --- ## Preamble ``` MIP40c3-SP69: Adding Proprietary High Frequency Trading Core Unit Budget (HFT-001) Author(s): @KenGriffin Contributors: @RoaringKitty @vbuterin Tags: hft Status: RFC Date Applied: <yyyy-mm-dd> Date Ratified: <yyyy-mm-dd> Budget Start Date: 2022-04-01 Budget End Date: 2022-12-31 ``` ## Sentence Summary This Budget Proposal for DAI expenses aims to fund the activities of the HFT-001 Core Unit as described in MIP39-SP42 Adding Proprietary High Frequency Trading Core Unit. ## Paragraph Summary This Budget Proposal for DAI expenses aims to fund the activities of the HFT-001 Core Unit as described in MIP39-SP42 Adding Proprietary High Frequency Trading Core Unit. This budget is necessary to provide a runway for the first nine months of HFT-001 activities. Following which, HFT-001 will finance continuing operations entirely through Core Unit profits and wind down the excess budget. ## Specification ### Motivation HFT-001 seeks to aggressively invest MakerDAO deposits and flashloans through MEV trading and off-chain HFT operations. The runway is necessary to finance the first few months of trading and compensation for Core Unit traders. During the Budget Term, the Core Unit will scale up trading algorithms until it can generate enough profits to sustain itself indefinitely through aggressive management of Vault deposits and through active DAI flashloan deployment. This will substantially increase earnings per MKR token on a balanced risk-adjusted basis. ### Core Unit ID HFT-001 ### List of Budget Implementations The HFT-001 budget is designed with the following in mind: * Paying for the operational costs to run the core unit for the first nine months * Having a buffer for unexpected legal, technical, or financial problems Therefore, a vote to ratify this MIP means MKR holders make a commitment to: A continuous funding model similar to those proposed by RISK-001, SES-001, and DUX-001. If any modifications are needed, they would be done on a six-month basis through “Modify Core Unit Budget Subproposal Process” (MIP40c3 subproposals.) If modifications are not proposed, the monthly payment amount gets rolled over from the initial proposal, if our work meets community and MKR holders expectations and demands. ### List of Budget Breakdowns ### Reporting Transparency *List of expenses by type, per the [Chart of Accounts](https://docs.google.com/spreadsheets/d/1qB1MIPCkZfdkuHSNk1udhscvTjg-mIBo/edit#gid=1506071077) and [Core Unit Operating Expense Budget Template](https://xyz)* |Expense Category|Total Budget Term|Monthly Equivalent|Monthly Expense Annualized (x12)| | --- | --- | --- | --- | |Compensation & Benefits|270,000|30,000|360,000| |Bonus|-|-|-| |Travel & Entertainment|-|-|-| |**Headcount Expense Subtotal**|**270,000**|**30,000**|**360,000**| ||||| |Admin Expense|450,000|50,000|600,000| |Freight & Duties|135,000|15,000|180,000| |Gas Expense|90,000|10,000|120,000| |Governance Programs|-|-|-| |Hardware Expense|5,625,000|625,000|7,500,000| |Marketing|-|-|-| |Professional Services|9,000|1,000|12,000| |Software Development Expense|4,500,000|500,000|6,000,000| |Software Expense|90,000|10,000|120,000| |Supplies|18,000|2,000|24,000| |Training Expense|-|-|-| |**Non-Headcount Expenses Subtotal**|**10,917,000**|**1,213,000**|**14,556,000**| ||||| |*10% Contingency*|1,118,700|124,300|1,491,600| |**Grand Total**|**12,305,700**|**1,367,300**|**16,407,600**| ### Budget Term The HFT-001 Budget will begin on April 1st, 2022 and run for up to nine months through to December 31st, 2022. ### Increasing Value to Maker Tokenholders Investing 12.3m DAI in setup costs for the HFT-001 Core Unit is a judicious use of Maker capital. We expect our delta-neutral high-frequency market making strategies and MEV extraction to recover the budget within 6 months after the end of the Budget Term. After this, we will return all excess profits after trader compensation and operating expenses to the DAO and to token holders. HFT-001 is a one-sided bet on at least a 50% increase in earnings per MKR in 6 months with capped downside of 12.3m DAI over the course of 9mos. ### Roadmap and Objectives 01-Apr-2022: Budget deployment 31-May-2022: Infrastructure deployed and underwater cables rolled out between London and Calais 30-Jun-2022: First strategy backtesting completed 31-Oct-2022: First go-live under real market conditions 31-Nov-2022: Full team of traders hired 31-Dec-2022: Return excess funds to DAO and winding down HFT-001 Budget ### Headcount Planning / Team Summary |Month|FTEs|Forecasted Total HC Expense| | --- | --- | --- | |Apr-2022|1|10,000| |May-2022|1|10,000| |Jun-2022|2.5|25,000| |Jul-2022|2.5|25,000| |Aug-2022|2.5|25,000| |Sep-2022|2.5|25,000| |Oct-2022|3|30,000| |Nov-2022|3|30,000| |Dec-2022|3|30,000| ### Contingency Planning |Expense Category|Base Case Monthly|Bear Monthly (10%)|Superbear Monthly (20%)| | --- | --- | --- | --- | |Compensation & Benefits|30,000|20,000|15,000| |Bonus|-|-|-| |Travel & Entertainment|-|-|-| |**Headcount Expense Subtotal**|**30,000**|**20,000**|**15,000**| ||||| |Admin Expense|50,000|45,000|40,000| |Freight & Duties|15,000|13,500|12,000| |Gas Expense|10,000|9,000|8,000| |Governance Programs|-|-|-| |Hardware Expense|625,000|500,000|200,000| |Marketing|-|-|-| |Professional Services|1,000|-|-| |Software Development Expense|500,000|500,000|500,000| |Software Expense|10,000|10,000|10,000| |Supplies|2,000|2,000|2,000| |Training Expense|-|-|-| |**Non-Headcount Expenses Subtotal**|**1,213,000**|**1,079,500**|**772,000**| ||||| |*10% Contingency*|124,300|109,950|-| |**Grand Total**|**1,367,300**|**1,209,450**|**787,000**| * Bear Case: Our bear scenario if market conditions turn and DAO profits are compromised. Our budget could function with 11% less runway. In this eventuality, we will scale back our FTE timetable and reduce admin and freight expenses accordingly. However, we will not be able to scale back any of the software development expenses as they are necessary to start our operation. * Superbear Case: Our superbear scenario. The bare minimum for executing our plans is with a budget 42% smaller than the base case. The FTE plans would be significantly scaled back and we would run our algorithms on a network-cluster of Pixel Chromebooks instead of high performance ASICs and on-site servers. This would push back our ROI timetable well past 6 months from the end of the Budget Term but the original timetable could still hold. * We are sufficiently aggressive in our base case that we do not feel there are any untapped opportunities worth stretching the budget over ### Unwinding and Defunding * Most of the budget will go to upfront investments in hardware equipment and tooling. In the event of an unwinding, the DAO would be able to judge whether to liquidate the stock of equipment or to keep it to repurpose for other Core Units * In either case, there will be a 100,000-200,000 DAI expense just to unwind all the hardware that has been set up and to move it to a new location for a new purpose * Any remaining budget in the operational wallet would be returned through DSSBlow. ### Proposal Parameters * **Minimum Feedback Period**: 1 Month. * **Minimum Frozen Period**: 1 Week. * Ratification Polls for MIP40c3 subproposals involving a Core Unit Budget Modification will have different requirements to MIP51c2. Instead of binary voting, they must use ranked-choice voting, with the options as follows: * A: Approve existing budget only (Business as usual) * B: Approve increased budget * C: Reject budget * D: Abstain In order for this Ratification Poll to conclude successfully and the contained proposal(s) move to Accepted status, each of the following conditions must be true: * `A` or `B` vote-weight must exceed `C` vote-weight when the poll closes. * Winning vote-weight must exceed the `Minimum Positive Participation` value of 10,000 MKR when the poll closes. [/quote]