Benchmark
A reference index or ETF — the S&P 500, KOSPI, a broad-market fund — that you measure your own performance against to tell whether a result is actually any good.
Formula
excess return = my return − benchmark return (also called active return, or alpha)
You can't judge your own return in a vacuum. An 8% gain sounds great until you learn the S&P 500 rose 15% over the same stretch — at which point you actually trailed the market. A benchmark is the yardstick you hold your results up against, usually a headline index like the S&P 500 or KOSPI, or a fund that owns the whole market such as VOO or QQQ.
The real point of a benchmark is to compare you against simply buying the index and sitting still. All the effort of picking names and trading in and out only pays off if it beats a do-nothing index position. How far you come out ahead is your excess return, or alpha. For the comparison to be fair, the benchmark has to match what you actually own — hold US large caps and the S&P 500 is the right rival; hold Korean stocks and it's the KOSPI.
The same logic carries over to calling direction. LDBD runs baseline bots that predict 'up' on VOO and QQQ every single time and treats them as the benchmark to clear. Major indexes have drifted upward over the long run, so blindly predicting up lands correct surprisingly often. Because every prediction is settled against real price data, the only way to know your skill is genuine — and not just the market carrying you — is to check whether you outscore those always-up bots.
Example
If my portfolio returns +18% in a year while the S&P 500 returns +15%, my excess return is 18 − 15 = +3 percentage points — I beat the market by three points.
How LDBD uses it
LDBD runs always-up baseline bots on VOO and QQQ as its benchmark. These bots are made visible on the leaderboard immediately — their tier/visibility threshold is a weighted resolved count of just 1, versus 5 for a regular predictor — so they always sit on the board as a reference floor that everyone's rate can be read against. And in the Elo-style Skill Rating, each prediction's expected value is the asset's historical up-probability, base_rate_up — meaning every call is scored against a 'market baseline' benchmark rather than against nothing.
FAQ
Why compare against just buying the index?
Buying a broad index fund and holding it earns the market's average return with zero analysis. If your stock-picking and trading can't beat that, there's little reason to bother — which is why a plain index position is the natural baseline to measure against.
What does beating the benchmark mean on LDBD?
LDBD's benchmark is a set of always-up bots that predict 'up' on VOO and QQQ. Because the market drifts upward, those bots are right fairly often, so you only demonstrate real skill once your leaderboard rate clears theirs instead of merely riding the same market.