One honest view of the book
Open spot positions, staged adds and the harvest queue sit together, so exposure can be read at a glance instead of assembled across scattered tabs.
Horizon
A private Lab trial in algorithmic spot trading.
Horizon accumulates spot coins inside hard gates, harvests when the plan allows, and protects free capital before chasing growth. Live operations now run from our bespoke Trading Management Dashboard — a custom management interface built in-house for this book alone.
Operations moved onto a bespoke Trading Management Dashboard
The biggest shift of the trial so far is not a strategy tweak — it is where the work happens. Day-to-day live operations now run from a bespoke Trading Management Dashboard: a custom management interface built in-house for this book alone, covering spot positions, harvest queues, capital health and every risk dial on one deliberate surface. It is private and unlinked, and it replaced the spreadsheet as the control plane.
Every dial, grouped by purpose
Entry caps, capital floor and harvest thresholds sit in categorised cards — capital, controls and risk — each folding away on its own, with every change logged against the cycle that picked it up.
Harvest and cycle telemetry
Each cycle, fill and harvest lands with a timestamp and a reason code, so maker-fill quality and pacing stay auditable long after the run has finished.
Guards visible, not implied
Capital-guard state, exposure limits and pause flags are surfaced on the interface, so a defensive posture is obvious rather than buried in code.
Since the desk landed it has grown the parts that keep it honest. The interface is now categorised into clearly grouped cards, so capital health, controls and risk limits each have their own space. Dial changes are made behind a PIN: the backend verifies it, and the desk only redraws once the write has been accepted. The ledger carries hand-added deposits and withdrawals beside the rows synced from the trading stack — each accepted only after the same PIN check — so cash flows can be reconciled for tax rather than reconstructed from memory. And the performance view is multi-timescale: seven-day, thirty-day, one-year, two-year and all-time ranges, with equity and unrealised P&L kept on separate axes.
Everything else on the desk stays read-only. There are no trade, cancel or order controls — the interface can change a dial or log a cash movement, and nothing more.
The dashboard is private to this Lab book. It is described here for context only — no address, credentials or account details are published, and no balances or P&L figures appear on this page.
Live market insights
Showing sample market data. The live feed is not connected yet.
Market context the dashboard surfaces while the book runs: which way the tape is leaning, how many spot names are actually being traded, how activity is split, and where the sharpest twenty-four-hour moves sit. Everything below is market prices and percentages — never balances, never profit.
Active market trend
Mildly bearish
Breadth has slipped below its thirty-day midline, so the book keeps accumulating inside its gates instead of stretching for size.
Total coins traded
27
Spot names live in the book — 2 opened and 1 trimmed this week.
Directional split
65%35%
Long accumulation bias against a smaller short-side hedge. Read as a share of trading activity, not of capital deployed.
Biggest gainers24 hours
Up to +8.4%| Asset | From | Now | 24h |
|---|---|---|---|
| SolanaSOL | $138.20 | $149.85 | +8.4% |
| InjectiveINJ | $21.40 | $22.96 | +7.3% |
| ArbitrumARB | $0.62 | $0.66 | +6.5% |
| RenderRNDR | $4.85 | $5.13 | +5.8% |
Largest drawdowns24 hours
Down to -6.1%| Asset | From | Now | 24h |
|---|---|---|---|
| CardanoADA | $0.58 | $0.54 | -6.1% |
| PolkadotDOT | $4.20 | $3.99 | -5.0% |
| CosmosATOM | $4.72 | $4.51 | -4.4% |
| AlgorandALGO | $0.21 | $0.20 | -3.9% |
Sample data · percentages only · not financial advice. From and Now figures are market reference prices in USD held for context beside each percentage, not account values — no balance or profit figure is published here.Trend, breadth and split read from the dashboard · snapshot
Timeline
Early build work began in January 2026. The formal twelve-month trial opened on 1 August 2026 and runs through July 2027. Below is the arc — not every deploy, just the shape of the journey.
Foundations
First automation loops, Sheets as a living desk, and the habit of measuring harvest quality rather than noise. Plenty of iteration before the formal trial clock started.
Twelve-month trial opens
A deliberate year of live spot-only learning: same personal accounts, clearer rules, and a promise to prefer resilience over spectacle.
Hard lessons from drawdowns
Volatile stretches exposed unbounded accumulation and thin free capital. That forced entry caps, capital guards, and a slower, more honest scoreboard.
Lean dual-Lambda stack
Main cycle plus a dedicated market-data worker — Python 3.13 on ARM64 — paced by EventBridge Scheduler, with DynamoDB locking so runs never overlap, and Secrets Manager for credentials.
Operations moved to the bespoke dashboard
Day-to-day control left the spreadsheet for the bespoke Trading Management Dashboard — one custom management interface for dials, positions, harvest queues and capital health. See what it does.
A categorised desk, protected writes and two-year history
- Categorised interface: dials and risk parameters are now grouped into distinct capital, controls and risk cards, so each part of the book reads at a glance.
- PIN-protected writes: a dial change, or a hand-added ledger entry, is accepted only once the backend has verified the PIN — the desk redraws after that, never before.
- Tax-tracking ledger and history: manual deposits and withdrawals sit beside the synced rows so cash flows can be reconciled, and performance charts run from a week to two years.
One owner for the cross-account sweep rule
The sweep rule that decides when profit leaves the book now reads its balance and its freshness from the bespoke Trading Management Dashboard's own backend rather than from a spreadsheet cell. The desk records a heartbeat and the book balance as it runs; the separate allocator gates on that record, holding a dual-read window beside the sheet before it cuts over. The spreadsheet stays the desk's display surface — not the trigger.
Steady harvesting under fixed dials
Live spot soak with frozen risk dials: maker-first harvests, sequenced adds, and the capital guard armed — all watched through the bespoke dashboard. Reading capital-growth quality day by day rather than inventing new gates every week.
What changed
A lot of machinery came and went between January and the trial. The lasting shifts are few and deliberate:
Scoreboard honesty
Focus moved from fill notional to daily grow and capital health — whether free capital can still breathe.
Entry caps (8×)
Hard caps versus entry size quarantine bleeding names and stop infinite accumulation on the same spot coin.
Capital guard (~35% floor)
If available capital as a share of the book thins too far, new entries pause until the book can recover room to manoeuvre.
Maker-first harvesting
Unlock harvests prefer resting limit fills — typically around a ~99% maker-fill ratio among planned harvest events.
Dashboard as control plane
Sheets carried the early build. Day-to-day dials, positions, harvest queues and capital health now live on the bespoke Trading Management Dashboard, while AWS carries the compute.
Writes with a receipt
The only writes the desk can make are a dial toggle and a manual ledger entry, each one PIN-gated and confirmed before the interface redraws. Cash movements stay logged for tax, and nothing on the desk can place, amend or cancel an order.
Highs, lows, and the growth showing now
High — the stack got quiet
Lean dual-Lambda pacing, a real mutex, and a Sheets desk that explains each cycle. Less drama in the infrastructure means more attention for the book.
High — harvest discipline
Staged unlocks with a min-profit floor and maker preference turned “close something” into a repeatable habit instead of a scramble.
High — growth under soak
With dials frozen, daily grow has been printing more consistently: free capital often sits comfortably above the guard floor (recently around the 80%+ available band), while harvests keep banking without inventing new risk switches.
Low — unbounded adds hurt
Early on, averaging without a hard entry cap tied up capital in spot names that would not recover quickly. That lesson is why the 8× cap stays locked.
Low — one bad kill-switch day
An aggressive fast-eject crystallised a painful cluster of closes. The soak response was simple: leave that live kill-switch off, keep the slower protective sweep armed, and prefer shadow telemetry before live ejects.
- 8×entry cap
- ~35%capital floor
- ~99%maker harvests
- 80%+available capital in recent soak
- 65/35long / short activity split
Lessons so far
Protect the floor before you feed
New names and adds only matter if free capital survives. Capital guard and name caps beat clever entries.
Fewer dials, longer soaks
Changing risk switches every bad afternoon destroys signal. Freeze a small set, measure capital growth, then decide.
Shadow before crystallise
Telemetry on deep-red names teaches recovery behaviour. Live mass-closes teach regret.
Architecture is a risk control
Non-overlapping cycles, secrets out of the sheet, and a desk you can read under stress matter as much as the signal logic.
What I may try next
Still inside the August 2026 – July 2027 window. Nothing below is a promise — only a handful of careful experiments once the current soak has said enough:
Tighter capital discipline after feeding slows
Only once the book stops growing names — refine the auto-deposit / available-capital posture without inventing new skip gates.
Selective recycling of stuck stubs
If evidence supports it, carefully free trapped capital on post-trial opens only — never a mass close of the deep-red book.
Leaner operator surface
Keep the dozen dials that matter; merge or freeze the rest so the bespoke dashboard stays human-scale.
Under the Lab
Built by Phil Stone
Private Horizon work in Northern Ireland. Meet Phil · Get in touch
Core automation sits on a lean AWS serverless stack: dual Lambda (Python 3.13 on ARM64), EventBridge Scheduler, DynamoDB for hot state and locking, Secrets Manager for credentials, and a data layer behind the bespoke Trading Management Dashboard. Spot trading only, and strategy detail and results stay private.
Lab boundary: Horizon is personal research for Phil Stone’s own accounts, trading spot assets only. Trading spot assets still involves risk, including the possible loss of capital. This page is an introduction only. It is not financial advice, not an invitation to invest, and not a product for clients.