Arena Blog

"Stocks have recovered from every war" is true — for one country. After the April 1940 high the Dow needed four years and nine months to get back to even, in 1914 the exchange was closed for four and a half months, and St Petersburg, Vienna, Tokyo and Shanghai are missing from the statistics because their exchanges ceased to exist. Across 39 markets, a diversified investor is down in real terms after 30 years in 12 % of cases.

Across 18 midterm years since 1954 the S&P fell 17.7 percent from its high on average — but only 5 of them had a 19-percent drop inside the second half. What the number measures, how much of it reaches Bitcoin (through the S&P, not through yields), and why Bitcoin's own midterm-year bears come from the halving.

A network's value grows with the square of its participants — says Metcalfe's Law. Where it came from, where it was ever measured, why it describes 15 years of Bitcoin well, and why the "seven-year low in active addresses" headline still counts no users.

The most detailed factsheet among three providers checked lists hit rate, best trade and worst trade — but not how far the portfolio fell or for how long. Five questions that make a return readable.

Since 1 July 2026 every crypto platform serving the EU needs a licence. For you as a user that means no ban, but this: without a licence you lose protections — segregated assets, liability, a complaints route. Which platforms are in the register, what happens to DeFi, USDT and your own wallet, and how to check a venue.

TradingView's chart copilot can connect to your own MCP servers. Tested on the Backtesting Arena connector: it answered market questions correctly and invented instructions about its own settings four times. Why that happens, and how to tell which answers to check.

A backtest has more parameters than the strategy description lists. The start day of a 3-day candle grid alone moves annual return by 6.66 points on average. What that means for every interval comparison, and the three questions that make the effect visible.

The Richmond Fed tested 30 years of income data against a strict definition: a K appeared in 2002–06 and 2010–13, not after COVID. Why short stretches still last, and why no average can show one.

A leveraged long grid was liquidated in 1 of 70 ninety-day windows since 2017 — because a grid never deploys more than half its notional. What leverage does instead: it deepens every drawdown and pays funding in the down windows.

The 25 percent tipping point quoted in every adoption argument comes from a naming game with 18 to 30 players per group and no right answer. It counts people; prices respond to dollars. What tips in large portfolios is not understanding but the answer to one question: what happens to me if I am the only one at zero?

A grid bot had the flatter drawdown in 138 of 138 windows since 2017, ranges set in advance. Not because it trades well: because it never invests more than half the money. What that protection costs in rally windows, and what a hindsight range was worth.

Four findings you can trade on — win rates, timeframes, regime filters, costs — and four questions that expose any performance claim. From 100,000 systematic backtests.