First, the answer: neither of them is true.
This time, it’s not about someone coming to pick you up, nor is it about someone digging a pit waiting for you to jump.
It’s a reasonable price for gold, and someone recalculated it again.
I. Why “picking someone up” doesn’t hold
“Picking someone up” has a prerequisite.
It assumes the valuation hasn’t changed—just that prices temporarily step aside so that those who missed the move can come up.
This time, the valuation has changed.
The U.S. 10-year Treasury’s real yield—i.e., the return investors truly get after subtracting inflation—stands at 2.90%.
This is the highest since November 2008.
Gold doesn’t generate interest.
If you hold gold, the only thing you give up is this risk-free real return.
Now that “abandoned stuff” is sitting at the highest level since 2008.
The backdrop is that in September, the Fed raised the policy rate by 25 basis points to 3.75%–4.00%, the first rate hike since 2023.
So gold doesn’t need any new bad news—its fair price just has to move lower on its own.
This isn’t a change of position—it’s a re-computation.
II. Why “digging a pit” doesn’t hold
Digging a pit needs a fake catalyst. If the main force smashes the market and the fundamentals don’t cooperate, that’s what we call a pit.
This time, the catalyst is real.
Over the past week, the nominal 10-year yield rose by 28 basis points: 27 basis points came from real yields, while the inflation expectations implied by the market moved only 1 basis point.
That means what has been pressured down this week has almost nothing to do with inflation.
And it’s not just the U.S. The intraday high of the 30-year U.S. Treasury reached 5.581%, the highest since 2002.
Germany’s 10-year government bond yield is 3.64%, the highest since 2009; the UK’s 10-year is 5.42%.
Sovereign bonds around the world were re-priced at the same time.
This isn’t a pit dug by whoever—it’s a global duration re-pricing.
III. The real evidence is at 4,288
On the morning of September 28 at 5:00, the intraday high touched 4,288.39.
Then at 6 o’clock it breaks the level straight away, and by 22:00 that evening the low hits 4,116.60.
Within one day, from the high to the low, the deepest drop was 4.01%.
The key isn’t this number—it’s the 4,288 level.
Over four days, the intraday highs landed a dozen times around 4,288, and not once did they manage to stand above it.
A level that has been repeatedly pushed up for four days but can’t be broken—once it breaks, it means the hands holding the chips at that price have been rotated.
Who it’s been handed to—see below.
IV. Today, what you should watch most isn’t the price, but the leverage.
For Binance XAU/USDT perpetuals, the open interest—i.e., the total quantity of positions not yet closed on this contract—was 109,954 contracts at 23:00 on September 27.
On the night of September 28 at 23:00, 172,833 contracts—up 57.2% over the prior 24 hours.
Even more critical is what comes after: after the crash, it didn’t come down.
Tonight at 21:00, there were 177,368 contracts—2.6% more than at 23:00 on the day of the crash.
Prices are falling, leverage is rising.
This shows that today isn’t long-side capitulation and exiting—if it were exiting, open interest should have dropped.
The funding rate has also never turned negative—it’s still at +0.009%.
The funding rate is positive—meaning longs are paying money to shorts.
Price falls, leverage rises—longs are still paying money to the shorts.
This combination in the futures market has a not-so-nice name.
Catching the order book—being the ones taking the other side.
V. Who is catching?
I won’t hide one piece of contrary evidence.
For large investors by open interest, the long share hit a low of 69.35%, and has now come back to 73.18%, rebounding by 3.8 percentage points.
This is smart money adding long positions again at lower levels, with a bias to the upside.
But that same day, the other set of numbers was the opposite: by account count, the share of all long accounts rose from 78.21% to 83.42%.
After big players reduced once, they then replenished; meanwhile retail kept adding longs throughout the entire selloff.
“Catching the falling” needs a主体 to catch. In the current setup, retail investors are catching while big players are picking up at lower levels.
This isn’t “catching the falling knives”—this is “turnover.”
VI. Structurally, where is it?
Now it’s 4,164.
The 20-day moving average is at 4,308, and the price is below all three moving averages.
In the past month or more of that range, the lowest closing level was 4,258, and it has already been broken down effectively.
Further down, between 4,028 and 3,948 is a vacuum zone.
So 4,116 isn’t a hard support—it’s just the bottom edge of the range you just broke down.
VII. When am I supposed to change my stance?
With three conditions—if any one of them appears, I’ll change my mind.
First, open interest falls back to around 120,000 contracts, or there’s a sharp drop in a single day.
Second, the funding rate turns negative and stays negative for several days.
Third, the 10-year real yield has been moving downward from 2.90%.
Before any of the three lines appears, treat the rebound as “repair” not “reversal.”
The U.S. August core PCE to be released tomorrow evening is the watershed; the expected year-over-year change is 3.4%.
VIII. So how do we pull the trigger?
If you’re using leverage, you can’t catch.
Right now it’s “a breakdown selloff before leverage has been cleared”—the most uncomfortable kind: it doesn’t give would-be bears who missed the move a cheap entry point, and it hasn’t washed the leverage clean.
For unlevered long-term positions built in batches, the 4,116 to 4,164 segment has structural support.
But you must treat re-testing 4,030—possibly even 3,950—as a normal scenario, not a tail event.
Finally,
Back to the headline: neither “catching the falling knives” nor “digging a pit.”
It’s as if the measuring rod for pricing has been swapped out for a new one.
If you focus on “who exactly is digging the pit for me,” you’ll keep doing the same thing over and over:
Before leverage gets fully cleared, someone has to take the last baton for others.
—MK守约
Data cut-off: 21:00 CST on September 29, 2026 (Beijing time). Gold price, open interest, open contracts, funding rate, and the long/short account ratio are all Binance XAU/USDT perpetual real-time data (24/7 trading; perpetual levels and COMEX futures levels must not be mixed). U.S. Treasury nominal and real yields are from the U.S. Treasury daily yield curve, latest as of September 28. Fed rate decisions come from official FOMC statements. The conditions and position discipline described in the article are for risk management and do not constitute investment advice.