Why Does the Quantity of a Token in My Portfolio Decrease?

In AI Rebalance strategies, users often notice two common phenomena — changes in token quantities and auto-invest failures. These situations are normal and directly linked to how the portfolio’s rebalancing logic works. Understanding this logic is essential to interpreting what’s happening with your holdings.

When you see that the quantity of a certain token has decreased, this is not a loss of assets, but rather a normal result of portfolio rebalancing.
The AI Portfolio continuously maintains the preset allocation ratio by selling assets that have increased beyond their target proportion and buying those that have fallen below it.
In other words, whenever one token’s quantity decreases, another token’s quantity must increase — the system is simply redistributing value within your portfolio to keep it balanced. This process ensures that your overall investment risk and return remain consistent with your original plan.

Let’s look at a detailed example.
Suppose you start with 1,200 USDT, allocated as 40 % ONE, 30 % BTC, and 30 % ETH.
At the beginning, your holdings are:

• ONE: 480 USDT (1 USDT per ONE → 480 ONE)
• BTC: 360 USDT (20,000 USDT per BTC → 0.018 BTC)
• ETH: 360 USDT (2,000 USDT per ETH → 0.18 ETH)

After some time, market fluctuations cause BTC to rise by 33.3 % (to 26,666.67 USDT) and ETH to fall by 33.3 % (to 1,333.33 USDT), while ONE remains unchanged.
Your holdings now look like this:

• ONE = 480 USDT (40 %)
• BTC = 0.018 × 26,666.67 = 480 USDT (40 %)
• ETH = 0.18 × 1,333.33 = 240 USDT (20 %)

BTC’s share has risen to 40 % ( 10 % above target), while ETH’s has dropped to 20 % (−10 % below target).
Assuming your rebalance threshold is 5 %, this deviation triggers an auto-rebalance.

The system’s goal is to restore your holdings to the original 40 % / 30 % / 30 % ratio.
It first calculates the excess and shortfall:

• BTC excess = 480 − (1,200 × 30 %) = 120 USDT
• ETH deficit = (1,200 × 30 %) − 240 = 120 USDT

The strategy then executes the following trades automatically:

• Sell BTC worth 120 USDT → 120 ÷ 26,666.67 ≈ 0.0045 BTC sold
  Remaining BTC = 0.018 − 0.0045 = 0.0135 BTC (≈ 360 USDT)
• Buy ETH worth 120 USDT → 120 ÷ 1,333.33 ≈ 0.09 ETH bought
  Remaining ETH = 0.18 0.09 = 0.27 ETH (≈ 360 USDT)

After rebalancing, your portfolio returns to the intended proportions:

• ONE = 480 USDT (40 %)
• BTC = 360 USDT (30 %) → quantity decreased
• ETH = 360 USDT (30 %) → quantity increased

This example shows that a decrease in token quantity is the natural outcome of rebalancing.
The system sells a portion of over-performing assets (high price, high weight) and reallocates that value to under-performing ones (low price, low weight).

This mechanism embodies the principle of “sell high, buy low”, allowing the AI Portfolio to maintain equilibrium, reduce risk concentration, and steadily accumulate arbitrage gains from market fluctuations — all automatically, without manual intervention.