Weekly Gold and Silver Update 18 July 2026

Executive Summary 

As of 18 July 2025, the three quantitative models present a nuanced picture: a near-term constructive signal from the GBDT forecaster is offset by persistent macro headwinds flagged by both the DXY-Metals Nexus Monitor and the Real Yields Monitor.

The GBDT model assigns a greater-than-60% probability of price gains for gold over the next 20 and 60 days, and above-50% probability for silver across both horizons. Median forecasts for both metals sit above current market prices. Three of four forecast horizons carry a positive skew — meaning the model sees more room to the upside than to the downside. A particularly notable observation this week is that the 20-day Q90 pinball loss scores are lower than both Q50 and Q75 for both gold and silver. As explained in the GBDT section below, this is a bullish leading indicator: it signals that actual prices have been consistently landing closer to the upper end of the model's predicted range, pointing to embedded upward momentum.

However, the macro backdrop remains challenging. The DXY-Metals Nexus Monitor composite signals for both gold and silver are firmly in the Strong Headwind zone, driven by a wide US-global rate differential, sustained dollar momentum, and the weight of oil-inflation-real yield dynamics. The Real Yields Monitor reinforces this — 10-year and 5-year TIPS real yields remain at elevated levels, raising the opportunity cost of holding non-yielding metals and suppressing the bullish case from the GBDT model.

Horizon	Q10	Q25	Q50 (Median)	Q75	Q90	Q95	Prob Up (%)
Gold 20-Day	$3,857	$3,932	$4,041	$4,108	$4,215	$4,504	63.7%
Gold 60-Day	$3,714	$3,899	$4,054	$4,194	$4,460	$4,864	63.7%
Silver 20-Day	$52.55	$54.58	$56.93	$61.52	$63.97	$70.27	58.1%
Silver 60-Day	$49.64	$52.54	$57.87	$63.69	$69.03	$81.55	60.2%

Skew Score Analysis

What is the Skew Score?

The Skew Score measures whether the model's forecast is tilted more toward the upside or the downside. It compares the gap between the optimistic forecast (Q90) and the median (Q50) against the gap between the median and the pessimistic forecast (Q10). A positive Skew Score means the upside potential is larger than the downside risk — a bullish lean. A negative score means there is more room to fall than to rise — a bearish lean. A score near zero indicates a broadly balanced outlook. The score is calculated as:

   Skew Score  =  (Q90 – Q50)  −  (Q50 – Q10)  ÷  (Q90 – Q10)

 

Horizon	Skew Score	Q90 – Q50 (Upside)	Q50 – Q10 (Downside)	Signal
Gold 20-Day	−0.0300	$173.23	$183.96	Neutral
Gold 60-Day	+0.0890	$406.50	$340.09	Bullish lean
Silver 20-Day	+0.2329	$7.04	$4.38	Bullish lean
Silver 60-Day	+0.1511	$11.16	$8.23	Bullish lean

 Pinball Loss — Forecast Accuracy and Calibration

What is Pinball Loss?

Pinball loss is a statistical report card for each individual quantile forecast. It measures how accurately each quantile — Q10, Q25, Q50, Q75, Q90 — has been tracking actual market prices over time. A lower pinball loss score for a given quantile means that quantile has been more accurately predicting where prices end up. Crucially, pinball loss does not tell you which direction prices are heading — it is purely a calibration measure, telling you which part of the model's forecast range is the best-calibrated against realised prices.

The scoring works asymmetrically by design. For a given quantile q, the model is penalised less when actual prices overshoot (come in above the forecast) and more when actual prices undershoot (come in below). This means that for Q90 to record a low pinball loss, actual prices must have been consistently landing near or above the Q90 line — in other words, the optimistic scenario has been the accurate one.

📈  Bullish Signal — Low Q90 Pinball Loss:  For both Gold and Silver at the 20-day horizon, the Q90 pinball loss is lower than both Q50 and Q75. This is a bullish leading indicator. When the high-end, optimistic quantile scores better than the middle quantiles, it means actual prices have been consistently landing closer to the upper end of the model's predicted range than to the middle. In plain terms: the model's most optimistic near-term call has been the most accurate one — a pattern historically associated with upward price momentum. Both metals exhibit this simultaneously, adding a notable bullish tilt to the near-term outlook.

DXY Monitor

This week's composite signals for both Gold and Silver are in the Strong Headwind zone. Several drivers are contributing to this reading:

The US-global rate differential remains wide in the dollar's favour. US 10-year yields continue to offer a premium relative to European and Japanese equivalents, supporting dollar demand and keeping capital flows tilted away from non-yielding metals. The DXY momentum panels confirm that the dollar has been trading above its medium-term moving averages, maintaining a structurally strong trend. The rolling correlation between the dollar and metals remains intact — there is currently no meaningful divergence signal that would suggest gold or silver are about to break free from dollar influence and rally on their own structural thesis. The oil-inflation-real yield interaction also leans unfavourable: energy prices and inflation expectations are not generating the kind of real yield compression that typically provides a tailwind for precious metals.

Real Yields Monitor

The Real Yields Monitor shows both 10-year and 5-year TIPS real yields at elevated levels this week. The 8-week rate-of-change panels — the key diagnostic layer in this model — remain in headwind territory, indicating that yields have not fallen far or fast enough in recent weeks to mark a genuine inflection. There is no signal of a sustained downward move in either maturity that would point to the early stages of a Fed pivot or an inflation re-acceleration.

The 5-year TIPS yield in particular is the one to watch. As the more forward-looking maturity, it is expected to lead any meaningful turn lower ahead of the 10-year. In other words, watch the 8-week delta to go below zero and form a green area.  At present, it has not provided that lead signal. Until DFII5 turns convincingly lower — ideally confirmed by the 10-year following — the fundamental case for pausing metal sales on the basis of real yield dynamics is not yet in place.