Backtesting - General Market

Backtesting shows what would have happened. It says nothing about what will happen. This is its only honest feature. The engine simulates portfolio performance from 2020 to today. Test a thesis, compare configurations via sweep mode, and deploy the result as a live Dex Traded Fund (DTF) — from the UI or API.

Past performance does not guarantee future results. The past is a story you tell yourself to justify a decision you have already made. Backtesting cannot account for future market conditions, liquidity changes, or black swan events. See Risks before committing capital.

Backtesting via the UI

Navigate to the Backtest section. The interface is simple. The questions it raises are not.

Select a Category

Choose an asset category to define the token universe. Layer 1s, DeFi Blue Chips, AI Tokens, Memecoins, Made in China, and 50+ more from CoinGecko and DefiLlama.Only coins listed on Bitget at some point are eligible. A backtest on untradeable assets is fiction, not simulation.

Configure Parameters

Parameter	Description	Example
Top N	Number of assets by market cap	`10`
Weighting	Allocation scheme (see below)	`mcap`
Rebalance	Days between rebalances	`30`
Fee	Simulated trading fee per rebalance	`0.1%`
Spread	Slippage multiplier	`1.0`

Run & Review

Click Run Backtest. Results include total return, Sharpe ratio, max drawdown, annualized return, a NAV chart, and final holdings. The numbers look definitive. They describe the past with perfect hindsight.

Backtesting via the API

All simulation endpoints are GET requests with query parameters. No authentication required. The data is public. The conclusions you draw from it are yours alone. Base URL: https://generalmarket.io/api

Run a Single Backtest

curl "https://generalmarket.io/api/sim/run?\
category_id=layer-1&\
top_n=10&\
weighting=mcap&\
rebalance_days=30&\
base_fee_pct=0.1&\
spread_multiplier=1.0"

Show Response

{
  "run_id": 4821,
  "config": {
    "category_id": "layer-1",
    "top_n": 10,
    "weighting": "mcap",
    "rebalance_days": 30,
    "base_fee_pct": 0.1,
    "spread_multiplier": 1.0
  },
  "stats": {
    "total_return_pct": 312.45,
    "annualized_return": 32.8,
    "max_drawdown_pct": -72.1,
    "sharpe_ratio": 0.94,
    "total_fees_pct": 1.8,
    "total_trades": 248,
    "total_rebalances": 72,
    "total_delistings": 5,
    "start_date": "2020-01-01",
    "end_date": "2026-02-27"
  },
  "nav_series": [
    { "date": "2020-01-01", "nav": 1.0 },
    { "date": "2020-01-02", "nav": 1.012 },
    "..."
  ],
  "cached": true,
  "computed_in_ms": 0
}

Query Parameters

Parameter	Type	Required	Default	Description
`category_id`	string	Yes	—	Category slug (from `/sim/categories`)
`top_n`	int	Yes	—	Number of top assets by market cap
`weighting`	string	Yes	—	Weighting scheme (see table below)
`rebalance_days`	int	Yes	—	Days between rebalances
`base_fee_pct`	float	No	`0.1`	Fee % per rebalance
`spread_multiplier`	float	No	`1.0`	Slippage multiplier
`start_date`	string	No	`2020-01-01`	ISO date (YYYY-MM-DD)
`force`	bool	No	`false`	Skip cache, force recompute
`threshold_pct`	float	No	—	Drift-based rebalance threshold (%)
`fng_mode`	string	No	—	Fear & Greed regime: `fear`, `greed`, or `both`
`fng_fear_threshold`	int	No	`25`	FNG index below this = fear
`fng_greed_threshold`	int	No	`75`	FNG index above this = greed
`fng_cash_pct`	float	No	`0.4`	% to move to stables during greed
`dom_mode`	string	No	—	Dominance regime: `btc`, `eth`, or `inverse`
`dom_lookback`	int	No	`30`	Lookback days for dominance trend
`vc_mode`	string	No	—	VC filter mode
`vc_investors`	string	No	—	Comma-separated investor names
`vc_min_amount_m`	float	No	—	Min funding round in $M
`vc_round_types`	string	No	—	Comma-separated round types

Weighting Strategies

Weighting	Query Value	Description
Equal	`equal`	1/N allocation across all assets
Market Cap	`mcap`	Proportional to market cap
Capped Market Cap	`capped_mcap_10`	Market cap with 10% single-asset cap
Square Root Market Cap	`sqrt_mcap`	Dampened market cap weighting
Momentum	`momentum_90`	Top by 90-day momentum score
Inverse Volatility	`invvol_60`	Lower volatility = higher weight
Dual Momentum	`dual_mom_180`	Combined momentum/volatility signal
Risk Parity	`risk_parity_60`	Equal risk contribution per asset
Min Variance	`min_var_60`	Minimize portfolio variance
Multi-Factor	`multi_factor_90`	Composite momentum + vol + mcap
Low Volatility	`low_vol_60`	Over-weight low-vol assets

DeFi-only strategies (for dl-* categories):

Weighting	Query Value	Description
TVL Weight	`tvl`	Proportional to Total Value Locked
TVL Capped	`tvl_capped_10`	TVL with 10% cap
TVL Sqrt	`tvl_sqrt`	Dampened TVL
Fees Weight	`fees`	Weight by protocol fee revenue
Revenue Weight	`revenue`	Weight by protocol revenue
Volume Weight	`volume`	Weight by trading volume
TVL Momentum	`tvl_momentum_90`	TVL trend-following
Fee Efficiency	`fee_efficiency`	Fees-to-TVL ratio
Yield Weight	`yield`	Weight by yield

Sweep Mode

Sweep mode tests many configurations at once. It varies one parameter and streams results for each variant. It is the fastest way to discover that most configurations perform similarly — and that the few that don’t are probably overfitting. Sweep dimensions: top_n, weighting, rebalance, category, threshold

curl -N "https://generalmarket.io/api/sim/sweep-stream?\
sweep=top_n&\
category_id=layer-1&\
weighting=mcap&\
rebalance_days=30"

Show SSE Event Stream

data: {"type":"progress","pct":14,"message":"Running top_n=5..."}

data: {"type":"result","run_id":4822,"config":{"category_id":"layer-1","top_n":5,"weighting":"mcap","rebalance_days":30},"stats":{"total_return_pct":245.3,"sharpe_ratio":0.87,"max_drawdown_pct":-68.2,...}}

data: {"type":"result","run_id":4823,"config":{"top_n":10,...},"stats":{...}}

data: {"type":"result","run_id":4824,"config":{"top_n":20,...},"stats":{...}}

Sweep	What It Varies	Values
`top_n`	Number of holdings	5, 10, 20, 30, 50, 100, 200
`weighting`	All weighting schemes	equal, mcap, momentum, invvol, etc.
`rebalance`	Rebalance frequency	14d, 30d, 60d, 90d, 180d + drift bands (3%, 5%, 10%, 15%)
`category`	Multiple categories	Pass `categories=layer-1,defi,ai`
`threshold`	Drift rebalance bands	3%, 5%, 10%, 15%, 20%

One call tests all variants. Results stream back as they complete. The temptation is to pick the best-performing result. The discipline is to ask why it performed best and whether that reason will persist.

Fetch Holdings from a Backtest

After running a backtest, fetch the portfolio composition. These are the holdings that survived:

curl "https://generalmarket.io/api/sim/holdings?run_id=4821"

Pass an optional date parameter (YYYY-MM-DD) to get holdings at a specific rebalance date.

Compare Benchmarks

Fetch BTC and ETH price series normalized to 1.0 at the start date. Every strategy must answer one question: did it beat holding BTC? Most do not.

curl "https://generalmarket.io/api/sim/benchmarks?start_date=2020-01-01"

Market Signal Integrations

Layer additional signals into any backtest. More signals do not mean more clarity. But they do mean more data:

Fear & Greed Index

Shift allocation between crypto and stablecoins based on market sentiment. Fear and greed are the only two emotions the market has. The index measures the ratio.

curl "https://generalmarket.io/api/sim/run?\
category_id=layer-1&top_n=10&weighting=mcap&rebalance_days=30&\
fng_mode=both&fng_fear_threshold=25&fng_greed_threshold=75&fng_cash_pct=0.4"

When FNG > 75 (greed), 40% of the portfolio moves to stables. When FNG < 25 (fear), fully deployed in crypto. The algorithm does what you cannot: sell into euphoria.

BTC/ETH Dominance

Adjust crypto allocation based on Bitcoin or Ethereum dominance trends.

curl "https://generalmarket.io/api/sim/run?\
category_id=layer-1&top_n=10&weighting=mcap&rebalance_days=30&\
dom_mode=btc&dom_lookback=30"

VC Funding Filter

Filter assets by venture capital backing. Smart money is not always smart. But it is always funded:

curl "https://generalmarket.io/api/sim/run?\
category_id=layer-1&top_n=10&weighting=mcap&rebalance_days=30&\
vc_mode=include&vc_investors=a16z,paradigm&vc_min_amount_m=10"

Understanding Results

Numbers do not lie. But they omit. The metrics below tell you what happened. They do not tell you what it felt like to live through it.

Key Metrics

Metric	What It Tells You
Total Return %	Cumulative gain/loss over the full period
Annualized Return	Return normalized to a yearly rate
Max Drawdown %	Largest peak-to-trough decline — worst-case pain
Sharpe Ratio	Risk-adjusted return. >1.0 is decent, >2.0 is strong
Total Fees %	Cumulative fees paid across all rebalances
Total Trades	Number of individual asset trades executed
Total Delistings	Assets removed mid-backtest (delisted from exchange)

What the NAV Series Represents

The nav_series starts at 1.0 and tracks portfolio value over time. A NAV of 3.12 means 212% return from the start date. A straight line on the chart. In reality, each day between 1.0 and 3.12 was a decision not to sell.

Deploy as DTF

A backtest that looks good enough becomes a live DTF with one click. The transition from simulation to reality is instant. The consequences are not.

Review Final Holdings

Check the holdings table. Confirm the weights match your thesis. The thesis looked good in simulation. It has not yet met the market.

Click Deploy Index

The button pre-fills the DTF creation form with the backtested assets and weights. One click from history to the present.

Confirm On-Chain

Add a name. Confirm the transaction. The strategy is now live. It belongs to reality now, not to the past.

The deployed DTF uses the weights from the final rebalance. It does not auto-rebalance. The backtest rebalanced for you. Reality will not.

Checklist: Running an Honest Backtest

Pick a Category That Matches Your Thesis

Do not backtest “all crypto” if you want a DeFi strategy. Narrower categories give cleaner signal. Broader categories give cleaner conscience.

Start with Equal Weight

Equal weight is the fairest baseline. If mcap-weighted cannot beat equal-weighted, concentration is hurting you. Equal weight is the null hypothesis. Beat it before you complicate things.

Use Realistic Fees

Default fee is 0.1%. If you are trading illiquid altcoins, bump spread_multiplier to 2.0+. Underestimating fees is the most common backtesting lie. It is also the most flattering.

Run a Sweep Before Committing

Use sweep=top_n and sweep=weighting to find optimal params. Do not cherry-pick the one configuration that looks best. That is not analysis. It is self-deception with a chart.

Check Drawdown, Not Just Return

A 500% return with -90% max drawdown means you would have watched 90% of your money disappear at some point. The return is theoretical. The drawdown was lived.

Look at Delistings

High total_delistings means the category contains flaky assets. The backtest handles this gracefully. Live trading does not.

Compare Against BTC

If your strategy cannot beat holding BTC on a risk-adjusted basis (Sharpe), the complexity adds nothing. BTC is the benchmark. Everything else is a bet against simplicity.

Documentation Index

​Backtesting via the UI

​Backtesting via the API

​Run a Single Backtest

​Query Parameters

​Weighting Strategies

​Sweep Mode

​Fetch Holdings from a Backtest

​Compare Benchmarks

​Market Signal Integrations

​Fear & Greed Index

​BTC/ETH Dominance

​VC Funding Filter

​Understanding Results

​Key Metrics

​What the NAV Series Represents

​Deploy as DTF

​Checklist: Running an Honest Backtest

Backtesting via the UI

Backtesting via the API

Run a Single Backtest

Query Parameters

Weighting Strategies

Sweep Mode

Fetch Holdings from a Backtest

Compare Benchmarks

Market Signal Integrations

Fear & Greed Index

BTC/ETH Dominance

VC Funding Filter

Understanding Results

Key Metrics

What the NAV Series Represents

Deploy as DTF

Checklist: Running an Honest Backtest