GLD, TLT, DBC, UUP Donchian ADX ATR Multi-Asset Long Only Backtest: Profitable, But Risk-Adjusted Evidence Is Weak

Executive Summary

Can a simple, volatility-scaled breakout model capture persistent trends across gold, bonds, commodities, and the U.S. dollar without relying on equity-market direction? This run covers 2010-01-01 to 2026-01-01 using quantstrat in R, with initial equity of $100,000. I compare it with buy-and-hold, apply base friction, and run a first sequence <span class="lf0-tooltip" tabindex="0" style="border-bottom: 1px dotted currentColor; cursor: help;" title="A deliberate check that makes assumptions harsher to see whether the result survives outside ideal conditions." data-tooltip="A deliberate check that makes assumptions harsher to see whether the result survives outside ideal conditions." aria-label="stress test: A deliberate check that makes assumptions harsher to see whether the result survives outside ideal conditions.">stress test</span>.

Research Card

The Setup

The counter-narrative here is that some ETFs don’t reward constant attention so much as they reward waiting for a clean break and then staying with it. My hypothesis is that a long-only trend filter can work across very different macro exposures-gold, duration, commodities, and the dollar-because each can spend long stretches in persistent regimes driven by policy cycles, inflation shocks, or flight-to-quality flows.

Results

The backtest only weakly supported my original hypothesis. Final equity was $118,877.74 on $100,000 start, with net P&L of $18,877.74 and CAGR of 1.09%, so the return story is modest at best.

Statistical Limits And Regime Dependence

The key question is whether the evidence survives sampling uncertainty, clustered trades, and regime shifts. P-values are diagnostics, not trading signals.

Regime dependency: this setup can degrade when volatility expands, gaps dominate, or the sector enters a persistent trend. The in-sample/out-of-sample split further down tests stability, not parameter-optimization skill.

Costs And Stress Test

I now run a simple friction pass before drafting the article. For this strategy test, the base case assumes 3.00 bps of <span class="lf0-tooltip" tabindex="0" style="border-bottom: 1px dotted currentColor; cursor: help;" title="The difference between the expected execution price and the actual fill. It usually gets worse in fast or illiquid markets." data-tooltip="The difference between the expected execution price and the actual fill. It usually gets worse in fast or illiquid markets." aria-label="slippage: The difference between the expected execution price and the actual fill. It usually gets worse in fast or illiquid markets.">slippage</span> per side and 1.00 bps of spread cost per side, or 8.00 bps per round trip. That is still a model, but it is a better starting point than pretending fills are free.

Watch the break-even round-trip cost: above that level, the historical edge is gone. Monte Carlo separates edge from path luck; the in/out row is a chronological closed-trade split (first 70% of trades vs. the rest), not walk-forward optimization. The out-of-sample half looks stronger here, but with only 13 in-sample and 6 out-of-sample trades this is far too small a sample to read as a real improvement — it mainly shows that this particular split isn’t a red flag, not that the edge is confirmed.

What Breaks It

The trade P&L distribution is ugly in the way that matters: most trades cluster near flat to slightly negative, while a few outsized winners carry the whole result. Across all 87 trades, the median trade was -$105.03 and the <span class="lf0-tooltip" tabindex="0" style="border-bottom: 1px dotted currentColor; cursor: help;" title="The percentage of trades that made money. It only matters alongside payoff size and loss size." data-tooltip="The percentage of trades that made money. It only matters alongside payoff size and loss size." aria-label="win rate: The percentage of trades that made money. It only matters alongside payoff size and loss size.">win rate</span> was only 38.9%, so the typical trade lost money rather than paying for the losers — the result rests on a small number of large winners (the largest single winner made $6,336) offsetting a majority of small losses.

Trade Log

Verdict

I set out to test whether a simple long-only breakout system, filtered by trend strength and regime, could extract durable upside from a small set of macro ETFs without relying on constant participation.

What would change my mind: the same rule needs to survive costs while closing the gap versus buy-and-hold, or prove useful as a timing overlay rather than a standalone strategy.

R Notes

The reproducibility contract is the saved artifact set plus the R scripts that created it. The key JSON/RDS files are recorded in the frontmatter.

# Rebuild the evidence layer for one saved strategy output
source('backtesting_blog_posts/statistical_diagnostics.R')
source('backtesting_blog_posts/walk_forward_diagnostics.R')
source('backtesting_blog_posts/analysis_test_selector.R')

out_dir <- file.path('backtesting_blog_posts', 'outputs', 'strat_donchian_adx_atr_multiasset_long_only_multiasset_donchian_adx_atr_trend_test')
per_trade <- readRDS(file.path(out_dir, 'per_trade_stats.rds'))
portfolio <- readRDS(file.path(out_dir, 'portfolio_summary.rds'))

diagnostics <- build_statistical_diagnostics(per_trade, portfolio)
walk_forward <- build_walk_forward_summary(per_trade)
test_plan <- build_analysis_test_plan(out_dir)

For the current <span class="lf0-tooltip" tabindex="0" style="border-bottom: 1px dotted currentColor; cursor: help;" title="Percentage Bands, a band-style indicator that marks stretched prices relative to recent behavior." data-tooltip="Percentage Bands, a band-style indicator that marks stretched prices relative to recent behavior." aria-label="PBands: Percentage Bands, a band-style indicator that marks stretched prices relative to recent behavior.">PBands</span> series, parameter_sensitivity uses a deterministic next-bar rerun helper because native quantstrat parameter clones can fail to propagate osMaxPos limits safely. Future strategy families should call register_parameter_sensitivity_builder() with their own sensitivity builder.

Important: Backtests are simulations. Past performance of a simulated strategy does not guarantee future results. Results do not yet fully account for real-world factors including <span class="lf0-tooltip" tabindex="0" style="border-bottom: 1px dotted currentColor; cursor: help;" title="The gap between the price buyers bid and sellers ask. Crossing that gap is a real trading cost." data-tooltip="The gap between the price buyers bid and sellers ask. Crossing that gap is a real trading cost." aria-label="bid-ask spread: The gap between the price buyers bid and sellers ask. Crossing that gap is a real trading cost.">bid-ask spread</span>, market impact, borrow costs, or taxes. This is for educational purposes only and is not investment advice.