AI bubble guide - EQOS Asia

Gold, cash and bonds in an AI bubble

Summarise this page with AI

The essentials in 30 seconds

  • Gold, cash, and bonds are buffers with different drivers than AI earnings, and each has its own risk.
  • Cash gives liquidity, bonds give income, and gold adds a price driver of its own.
  • Illustration: a 30% defensive sleeve split into 10% cash, 15% bonds, and 5% gold.
  • Deploy cash in 3 tranches after falls of 10%, 20%, and 30%.
  • The EQOS Asia AI study scores the enterprise AI market at 0.337 (April 2026).

When AI prices run ahead of earnings, many investors look to assets with other drivers: cash, bonds, and gold. Each plays a different role and carries a different risk. This guide, as of October 2026, sets out what each one does in a portfolio exposed to an AI bubble, how to size them, and how to use cash after a fall. It is one layer of how to hedge a portfolio against an AI bubble. This page is general information, not investment advice.

Compare the EQOS Asia offers and request a 30-minute pre-diagnostic.

See the offers

What roles do gold, cash, and bonds play in an AI bubble?

Each asset does one job. Cash gives liquidity: you can use it at once, at its full value. Bonds give income and a repayment date. Gold gives a price that follows its own drivers and depends on no company. None of these three earns its return from AI spending, and that independence is what a buffer offers.

Independence has limits. No asset in this group is guaranteed to rise when AI prices fall. The EQOS Asia AI study records valuations up ×4 in 18 months (April 2026), and a broad sell-off after such a run can touch many assets. Treat the three as shock absorbers that lower the size of the fall, and plan for the fall that remains.

How much does cash help a portfolio in an AI bubble?

Cash gives two benefits. It holds its nominal value when AI prices drop, and it gives you the means to buy after a fall. The second benefit is the larger one. A portfolio with cash can act when prices are low, while a fully invested portfolio must wait or sell something else.

Cash costs something too. Prices rise over time, so cash buys less each year, and it earns less than assets that grow. A cash share that is too high drags on returns in a long boom. Choose a share that matches your need: a reserve for living costs plus a smaller part for buying opportunities. Keep the two parts in separate accounts, so the opportunity part stays available for its purpose.

Which bonds fit a portfolio exposed to an AI bubble?

Government bonds pay interest and return the principal at maturity. Their prices move with interest rates: when rates rise, existing bonds fall, and long-dated bonds move more than short-dated ones. A short maturity lowers that swing. Pick the maturity by the date when you may need the money. A bond that matures on that date returns its principal, whatever the market did in between.

Corporate bonds add credit risk, the chance that the borrower fails to pay. Bonds issued by AI-linked borrowers carry some of the same risk as their shares. If bonds serve as a buffer, check who issued them. A bond from an issuer with no link to AI demand gives a cleaner break from the theme.

What can gold add to a portfolio during an AI bubble?

Gold pays no interest and has no issuer. Its price follows its own drivers, such as real interest rates, the dollar, and demand from buyers. That separate set of drivers is the reason to hold some. It adds variety to a portfolio whose other assets respond to similar forces.

Gold can also fall, sometimes at the same time as shares. It swings in price and earns nothing while you wait. Hold it for its different behaviour and keep the share modest. Many plans set a small share, and review it each quarter next to the AI weight. Storage, fees, and the gap between buying and selling prices add small costs, so compare the ways to hold it.

How much gold, cash, and bonds should an AI-heavy portfolio hold?

Illustration, not a recommendation: a defensive sleeve of 30% of the portfolio, split into 10% cash, 15% bonds, and 5% gold. The figure shows the job and the main risk of each. Set your own sleeve from two inputs: the loss you can carry if AI prices fall, and the cash you may need in the next 2 years.

Revisit the sleeve when your life changes or when the AI weight moves a lot. A larger AI weight calls for a larger buffer. A written split, with a review date, keeps the decision calm. The sleeve works with the rebalancing rules of this branch, which restore the split after prices move.

Job and main risk of cash, bonds and gold in an AI-heavy portfolio
Each defensive asset has one job and one main risk.

How do you put cash to work after an AI bubble falls?

Decide the plan before the fall. One approach uses 3 tranches: buy one third of the planned amount after a 10% fall in your AI target, another third after 20%, and the last third after 30%. Each step has a trigger, so no step depends on guessing the bottom.

The plan has a cost. If prices rise without falling, the cash stays idle, and you may buy later at a higher price. If prices fall further than 30%, you hold the last tranche at a loss for a while. A staged plan lowers regret on both sides, which helps you keep to it. Write the triggers and amounts in advance, and the plan runs on rules.

Three tranches to deploy cash after falls of 10%, 20% and 30%
A staged plan to use cash after a fall.

How does EQOS Asia scoring test a defensive mix in an AI bubble?

The EQOS Asia AI study scores the enterprise AI market at φ(S) 0.337, a fragile zone where a shock can tip the system, with bifurcation as the verdict. A defensive mix should help in a break and cost little in a soft landing. Scoring both cases shows what your buffers are worth.

The Λ simulator gives a first estimate of your case in 10 minutes. For larger portfolios in Singapore, ASEAN, or Hong Kong, a 30-minute pre-diagnostic with EQOS Asia applies the same scoring to your options and returns a ranked set of choices.

Frequently asked questions

Do gold, cash, and bonds protect against an AI bubble?

They act as buffers with drivers other than AI earnings. None is guaranteed to rise when AI prices fall, and each carries its own risk: inflation for cash, interest rates for bonds, and price swings for gold. Together they lower the size of a fall.

How much cash should I hold in an AI-heavy portfolio?

Hold a reserve for living costs, plus a smaller part for buying after a fall. The right share depends on the loss you can carry and the cash you may need in the next 2 years. A larger AI weight calls for a larger buffer.

Which bonds suit a buffer against an AI bubble?

Government bonds with a short maturity move less when rates change. Corporate bonds add credit risk, and bonds of AI-linked borrowers share part of the AI risk. Check the issuer and the maturity before you buy.

Is this advice on gold, cash, and bonds?

This page is general information, not investment advice. The split in the illustration is an example for learning. The EQOS Asia study leaves the timing of any fall open (EQOS Asia AI study, April 2026).

Test your knowledge

1. What does cash offer that other assets give less of?

  • Immediate use at full value
  • A dividend
  • A guaranteed rise
Show the answer

Immediate use at full value.

2. What moves the price of existing bonds when rates rise?

  • They rise
  • They stay fixed
  • They fall
Show the answer

They fall.

3. What is a reason to hold some gold?

  • It pays interest
  • It follows its own price drivers
  • It always rises in a sell-off
Show the answer

It follows its own price drivers.

4. In the illustration, how much of the defensive sleeve is gold?

  • 5%
  • 10%
  • 15%
Show the answer

5%.

5. When does the staged plan buy its second tranche?

  • After a 10% fall
  • After a 30% fall
  • After a 20% fall
Show the answer

After a 20% fall.

About the author

Portrait of Stéphane Jambu

Stéphane Jambu

President, EQOS Forecast Asia

Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile

Sources: EQOS Asia AI impact study (eqos-asia.com/case-studies/ia-2026.html), Λ simulator (eqos-asia.com/simulator.html) and contact page (eqos-asia.com/contact.html), consulted on 22 October 2026. Page updated on 22 October 2026.