Smart Rebalancing

Risk disclosure: The various rebalancing methods described in this chapter (band / smart / ratio / Beta-threshold), the backtested annualized returns, the four-market-condition backtest matrix, the monthly rebalancing timing, and so on, are all backtest summaries of public historical data and explorations of the mathematical models of asset allocation. Past backtested performance is no guarantee of the future; different backtesting methods and different parameter assumptions will yield different results. Nothing in this chapter constitutes individual investment advice or specific buy-and-sell timing instructions.

In the wild winds and towering waves of the capital markets, investors often find themselves caught in the crossfire of fear and greed. The pursuit of the ultimate long-term return is inevitably accompanied by violent short-term volatility, and especially once high-volatility leveraged instruments are added into a portfolio, risk control becomes the decisive judgment that determines an investment’s survival or ruin. Smart rebalancing is a protective mechanism of great mathematical discipline; it can transform an asset allocation into an autopilot system that ignores market noise. This is not merely a technique for locking in profit; it is the last line of defense that keeps assets from being shaken to pieces by volatility in an extreme bear market.

Before we get into the operational details of this chapter, we must first establish a principle of distinguishing the objects involved, so that readers do not conflate two separate strategies. The object of “hold and never sell” is the underlying index (such as QQQ / 00662) placed in a general account, and the purpose is to let it enjoy long-run compounding and to reduce transaction taxes and friction. The object of smart rebalancing, by contrast, is the dynamic adjustment between the leveraged fund (such as QLD / 00670L) and the cash position, and the purpose is to handle the volatility decay of leverage and survival in an extreme bear market. The two run in parallel without conflict: the underlying focuses on long-run compounding, rebalancing focuses on risk control, and losing either one leads to collapse.

The Operational Boundary Between Holding Forever and Rebalancing

Many investors feel puzzled: if we emphasize treating the index as a fixed asset to “hold and never sell,” why do we still need smart rebalancing? These two do not actually conflict; the key is that we must clearly divide the different account-operation rules and strategic boundaries between the “underlying fund” and the “leveraged fund / cash.”

“Hold and never sell” applies only to the underlying index fund (such as QQQ) placed in a general brokerage account, and its purpose is to avoid the capital-gains tax of frequent trading, letting assets compound without limit in the long river of time. Smart rebalancing, by contrast, is reserved exclusively for the dynamic adjustment between the leveraged fund (such as QLD), which carries extremely high volatility decay, and the cash position.

Many investors, upon hearing “I want to earn a bit more,” go off to research option techniques such as manually selling covered calls (Sell Call), imagining that holding QQQ can also collect option premiums as extra passive income. Teacher James has an extremely blunt rejection of this:

There is no need for this. If all of a sudden QQQ jumps 3% or 5% in a single week, doesn’t your QQQ get called away? Otherwise you lose money, or you have to buy it back at a loss. The call you originally sold for $2,000, you might have to spend $4,000 to buy it back. What is good about that? You should do 442, do not trade at all; that is far better than these petty little tricks like selling the call. (Video 00497)

The key issue lies in QQQ’s explosiveness. The NASDAQ-100 frequently posts violent surges of 3% to 5% in a single week, and when you sell a covered call for the sake of a few hundred dollars in premium, a sudden big rally leaves you with only two choices:

• One, let your precious QQQ position get called away (losing forever the core engine of unlimited future compounding).
• Two, buy it back to close out in the opposite direction (the contract you originally sold for $2,000 might cost $4,000 to buy back, a $2,000 loss before you have even earned the premium).

Adding it up, there is not only no gain, but the discipline of “hold and never sell” is broken. The stance of the CLEC system is very clear: rather than playing the petty little trick of selling calls to earn a meager rent, it is better to directly execute the 442 allocation and put capital truly into a leveraged position (QLD / 00670L) that can keep pace with QQQ’s full upside. A covered call is essentially “trading unlimited upside for a limited premium,” and this exchange ratio is entirely unfavorable to the long-term investor.

▲ Figure 12-1 Rebalancing is an internal conversion within your assets that locks in profit

A leveraged fund must undergo smart rebalancing. Do not skip it for the sake of chasing performance, or in the end it will be knocked back to square one overnight, and all your effort will have been in vain. (Video 00526)

If a leveraged fund runs into a long bear market and you do not pump water out to lock in profit, it will face the trap of bankruptcy. You can picture the funds as two cups of water; when the stock cup overflows because of a big rally, pour some of it into the cash cup. This principle runs in parallel without conflict: the underlying fund focuses on long-run compounding and low transaction friction, while the rebalancing system focuses on risk control of the leveraged position. Staying in the market is the sole prerequisite for compounding.

For investors who find percentages abstract and hard to grasp, we can switch to an even more intuitive visualization. Picture your total assets as a 100-cell block of Tetris. Suppose total assets are NT$10 million, each cell represents NT$100,000, and a 433 allocation is 40 cells of underlying + 30 cells of leverage + 30 cells of cash. When the leveraged position rises to 35 cells, the act of rebalancing is to “move the extra 5 blocks from the leverage column back to the cash column.” This image of “blocks moving house” is far easier for the brain to remember than abstract percentage arithmetic when it comes to the concrete action you should take each year, and it is an effective mental model for lowering the barrier to execution.

Execution frequency has an equally clear boundary: measured in years, not adjusted frequently within the day or the week. When the market rises, follow the rule and move 30% of the leveraged gains into cash; when the market falls, follow the rule and add to leverage. You must not add on impulse out of emotion, nor delay locking in profit on the excuse of “let’s wait a little for a higher point,” or your cash reservoir will be drained dry before the next bear market.

A slogan often heard in the community is “buy on every green” — buy as long as the day’s candle is green. This sounds proactive, but it is in fact a slogan severely lacking in standardization. A 1% drop is green, and a 10% drop is also green, so exactly how much should you buy? If the market falls for a week straight, buying “on every green” each day will drain the cash reservoir dry within a few days. The people who shout this slogan are usually those holding too large a cash position and too little stock exposure — they have never truly borne the discipline of asset allocation; they merely use “buy on every green” as an excuse to dump idle cash into the market. For investors who have already completed their asset allocation according to CLEC discipline, the true mission of the cash position is to serve as “ammunition for annual rebalancing” and as “an oxygen tank for a black-swan event,” and it should never be chronically consumed by a vague signal like “on every green.”

Many beginners, when carrying out rebalancing, always fret over tiny proportional errors, for example: “The underlying is now at 41% and leverage is at 39%, so shouldn’t I sell 1% to top it up?” This kind of thinking completely departs from the original intent of rebalancing. The underlying logic of rebalancing is not to keep the asset proportions frozen forever at a perfect 433 allocation like a robot, but to solve “the survival problem of high-volatility assets in an extreme climate.” Therefore, the supreme guiding principle in practice is: “handle only the extremes, ignore the minor ripples.”

Only when the market posts an annual-scale big rally or big crash, causing exposure to deviate severely off track, do you step in and intervene. During ordinary small swings of 5% to 10%, just let the proportions drift naturally. Overly frequent trading not only generates tax friction, but also interrupts the running momentum of assets in a bull market. Learning to catch the big and let go of the small is a necessary path to transforming from a retail investor into a capitalist.

To understand further why frequent rebalancing harms performance, we can dissect the mathematical mechanics of bull-and-bear cycles. If you adopt high-frequency automatic rebalancing such as “monthly / quarterly,” it will produce, under different market conditions, two behavior patterns harmful to long-term returns. In a long-rising bull market, automatic rebalancing forces the investor into “passively reducing positions against the trend” — it keeps selling and locking in the very assets that are surging strongly, interrupting the compounding momentum that could have kept running, and this is especially fatal for explosive targets like QQQ / QLD.

In a long-falling bear market, automatic rebalancing again forces the investor into “adding to positions against the trend and catching the falling knife” — it keeps pouring funds into the weak position that is falling continuously; although the tail end of the decline will rebound into a gain, the multiple add-ons in the middle of this stretch will severely drag down the overall cash level and increase paper losses. When you stretch the time span to cover a complete cycle of bull and bear, “the defensive advantage of high-frequency rebalancing” and “the drag disadvantage of high-frequency rebalancing” cancel each other out, and in the end the total return may even lag behind the simple version of “executing once a year.” This is precisely the mathematical basis for why the CLEC system strictly stipulates that “rebalancing frequency must be measured in years, and never adopt high-frequency automatic rebalancing” — make fewer decisions, and you actually earn a few more percentage points of compounding.

Four Ways of Operating Rebalancing

Rebalancing can be divided into four approaches by intervention frequency and rule, suited to different identities and risk preferences:

The Six Categories of Rebalancing and Their Strict Prohibitions

The practical drills of Chengfeng Linghang further distill rebalancing into six real-world scenarios, and investors must choose the correct tool according to their present fund situation; any confusion may lead to a performance disaster. (Chengfeng Linghang, “EP04”)

Rebalancing has one “bankruptcy taboo”: you absolutely must not delay locking in profit just because the market has not returned to its high. Looking back at 2000 to 2009, if in the 2003 rebound you felt “not back to breakeven yet” and did not replenish cash, then by the time the market fell again in 2007, you would have gone straight into bankruptcy by 2009. In addition, “pledging to the hilt and not rebalancing” is strictly prohibited; that is simply playing Russian roulette!

In practical operation, the criterion for judging an “up year” versus a “down year” must be clear: always compare only against “the same period last year,” and absolutely never against “the initial invested principal” or “the historical high.” For example, if 00670L falls sharply in the first year and rebounds in the second year but has not yet risen back to the first year’s level, the second year is still judged an “up year” and the profit-locking mechanism is triggered — because the comparison baseline is “the prior year” rather than “the initial principal.” This discipline breaks the retail investor’s most common mental-accounting myth: “it only counts as a profit once it rises back to cost price.” The market never cares about your cost price; it cares only about the present market value. By the same token, the trigger for rebalancing cares only about “the direction relative to last year,” and not about how much you lost in the past.

The Live Comparison Between Mindless and Smart Rebalancing

When discussing rebalancing, the “mindless rebalancing” of periodically distributing asset proportions evenly is often mentioned in common parlance (for example, keeping the ratio of underlying : leverage : cash/short-term bonds forever at 4:3:3). However, the core of investing is not merely to pursue the highest performance on paper; more importantly, it is to ensure “survivability” in an extreme environment.

Mindless rebalancing carries extremely high risk; you must adopt smart rebalancing, meaning that when the market rises you pull funds back quickly, and when the market falls you add in slowly. Historical backtests show that the core discussion should take the “433 allocation” (40% underlying, 30% leverage, 30% cash) as the unified baseline of this chapter. If you deviate to a low-cash version, then when a long bear market like 2000 to 2008 hits, the cash buffer is easily exhausted ahead of time.

Let us first look at the “433 weight structure” and the “up-year profit-locking mechanism,” the two figures corresponding respectively to the “433 static structure” and the “up-year profit-locking flow.”

▲ Figure 12-2 The 433 weight structure of the standard three-container allocation

▲ Figure 12-3 The up-year profit-locking mechanism, where overflow gains are shifted into cash

The key point when combining the “433 weight structure” with the “up-year profit-locking mechanism” is that the allocation code, the fund flow, and the execution rule must be consistent, or rebalancing degenerates into a mere slogan. By contrast, smart rebalancing adopts the asymmetric operation of “lock in 30% of profit in up years, replenish positions slowly in down years.”

This asymmetric design ensures that even in a decade-long great bear market, the cash reservoir still retains at least 10 to 15 years of living oxygen. Live historical testing shows that the combination of smart rebalancing and the income-recognition flywheel may lag behind mindless lump-sum investing during the first six years of stock accumulation, but in the seventh year it welcomes a “golden cross.” By the twentieth year, the assets of the practitioner of the system will be about 2.57 times higher than those of the linear grower!

Many people wonder: in an up year, should you lock in 30% of profit or 50%? The quantitative hard evidence indicates that the 30% profit-locking ratio, while balancing safety, retains the greatest compounding momentum. Especially during the “stock-accumulation phase” when you do not yet need to withdraw living expenses, keeping the selling ratio of the leveraged fund (QLD) in an up year down “between 10% and 30%” gives a final-outcome performance far superior to selling 40% or 50%.

The underlying mathematics is this: in a long bull market, the less you sell, the more high-explosiveness chips you retain, and the stronger the momentum for assets to sprint upward. In a long bear market, because the down-year replenishment draws a fixed amount from the cash reservoir to buy in, unrelated to how much you sold in the past, the defensive power is not thereby weakened. So during the stock-accumulation phase, letting profits run in the market as much as possible, and keeping the profit-locking sell ratio below 30%, is the best mathematical solution that balances stability and ultimate performance.

A concrete “vegetable seedling” metaphor can explain this more intuitively. If you use mindless rebalancing in a long bull, it is equivalent to harvesting in one shot all the “vegetable seedlings” (compounding chips) that the leveraged position has grown, forcibly pressing it back to the original proportion. The newly sprouted vegetables are all cut down before they have grown up, so naturally there can be no great harvest afterward. Smart rebalancing harvests only a portion (30%) as defensive replenishment, and the remaining seedlings stay in the field to grow to their fullest and keep compounding — this is precisely the core reason why smart rebalancing far outperforms mindless rebalancing in a long bull. A leveraged fund is itself a highly productive seedling field, and it absolutely must not be treated with “mechanized average harvesting.”

A concrete numerical example can show the calculation logic of mindless rebalancing more intuitively. Suppose an initial NT$1 million of assets adopts the 433 allocation:

• 00662: NT$400,000 (40%)
• 00670L: NT$300,000 (30%)
• 00865B: NT$300,000 (30%)

In an “up year,” the market values of the three positions change to NT$440,000 / NT$380,000 / NT$300,000, and total assets swell to NT$1.12 million.

Proportions after drift: 39% / 34% / 27%

The mindless-rebalancing approach is to combine leverage and cash-like assets and then split them evenly: 38 + 30 = NT$680,000, split evenly into two lots of NT$340,000. The three positions are adjusted to:

• 00662: kept at NT$440,000 (39%)
• 00670L: pulled back from NT$380,000 to NT$340,000 (30.5%)
• 00865B: topped up from NT$300,000 to NT$340,000 (30.5%)

Beta back to 1.0: 0.39 × 1 + 0.305 × 2 + 0.305 × 0 = 1.0

In a “down year,” the market values of the three positions change to NT$360,000 / NT$220,000 / NT$300,000, and total assets shrink to NT$880,000.

Proportions after drift: 41% / 25% / 34%

Handling it the same way: 22 + 30 = NT$520,000, split evenly into two lots of NT$260,000. The three positions are adjusted to:

• 00662: kept at NT$360,000 (41%)
• 00670L: topped up from NT$220,000 to NT$260,000 (29.5%)
• 00865B: pulled back from NT$300,000 to NT$260,000 (29.5%)

Beta back to 1.0 the same way: 0.41 × 1 + 0.295 × 2 + 0.295 × 0 = 1.0

Reading this calculation together with the seedling metaphor, you will understand why mindless rebalancing, though mathematically clean and returning Beta to normal, mechanically harvests the overwhelming majority of gains back in a long bull. In the up year, 00670L clearly earned NT$80,000, yet the mindless version immediately cuts NT$40,000, which is equivalent to forcibly zeroing out half the leveraged gains every year.

The Full Panorama of Smart Rebalancing in Chengfeng Linghang’s Four-Market-Condition Backtest

Chengfeng Linghang conducted a complete four-dimensional cross-backtest of “market rise/fall × allocation formation × rebalancing method × withdrawal scenario,” laying out the debate of “whether or not to rebalance, and how to do it” using real historical data. The backtest framework: 4 market conditions × 3 allocations (442/433/703) × 4 rebalancing methods (0-none / 1-smart / 2-mindless / 3-ratio) × 3 scenarios (retirement without withdrawal / retirement with 2% pledge-borrowing / accumulation-phase periodic investment), with the figures being annualized returns. (QLD / 00670L was only listed in 2006; the leveraged position for 1985 to 2006 is derived from a 2× leverage model synthesized from the NASDAQ-100 daily return × 2, not the actual ETF history.)

Extracting the most crucial comparison across the four market conditions (using the retirement with 2% pledge-borrowing scenario as the example):

Several key conclusions can be distilled from this matrix:

• “No rebalancing” has the best surface performance but conceals a death risk: over the long term, 442 with no rebalancing reaches an annualized 28.18%, but that is because the leveraged position is allowed to swell freely, its Beta eventually approaching 2, and in the one and only epic-scale stress test it triggers a blowup — if you retired at the 2000 peak with a 442 allocation + 2% pledge-borrowing + no rebalancing, by 2008 the maintenance ratio would fall to 166.56%, breaking below the 167% renewal death line. This is precisely why, even though its long-term performance looks beautiful, the CLEC system still strictly prohibits “no rebalancing.”
• The key warning for 442 in flat years: in most scenarios “smart rebalancing” outperforms “mindless rebalancing,” but the 442 allocation in flat years is a counterexample — smart rebalancing at -1.04% is far worse than mindless rebalancing at 3.25%. The reason is that in a flat, choppy market, the asymmetric action of “lock in on the rise, add to leverage on the fall” ends up doing both sides backward. If an investor adopts the 442 allocation, they must be mentally prepared that “flat-year performance will lag the mindless version.”
• 703 does not suit “ratio rebalancing”: because there is no leverage at the start, ratio rebalancing keeps suppressing performance; 703 must use smart rebalancing, letting the system gradually convert the cash position into leverage (2×) during the annual operation, so that performance is not nailed down.
• 433 is the most balanced choice for the retirement phase: across the four market conditions, 433 is never the “best,” but also never the “worst,” striking the optimal balance between volatility and performance. For the retirement phase, we recommend directly adopting 433 + smart rebalancing, which is also the underlying backtest support for the CLEC mainline rebalancing rule.

The final recommendation, split by group:

• Accumulation phase: 703 suits those who do not accept leverage; 433 for those pursuing ordinary performance; 442 for the aggressive who accept leveraged volatility.
• Retirement phase: 703 for those who cannot accept leverage; 433 as the recommended option; 442 only for those who have another source of living expenses and do not withdraw from their investment assets.

Backtest warning: All the annualized figures above are based on specific historical intervals and parameter assumptions; past performance does not represent the future, and different backtesting methods will yield different performance. The asset allocation and the rebalancing method themselves cannot directly determine performance; you must first understand the parameters, properties, and applicable scenarios behind each number.

Source: Chengfeng Linghang four-dimensional backtest matrix (market condition × allocation × rebalancing × scenario).

Chengfeng Linghang’s Backtest of the Annual Rebalancing Date

After choosing “1-smart rebalancing” as the annual operation, “which day to execute it” is another underrated detail. Chengfeng Linghang compiled statistics on the historical data of a total of 310 months from March 1999 to December 2024, analyzing each month’s “share of up/down months across the full period and the cumulative gain/loss”:

From the statistics, the two most noteworthy extreme months are:

• November: up months are 6.5% of the full period (that is, 20 up months, the most of the year), the strongest up month (that month’s own up rate is about 20 ÷ 26 ≈ 77%).
• September: down months are 4.5% of the full period (that is, 14 down months, the most of the year), with a cumulative decline of -100.72% (the deepest pit of the year), the weakest down month.
• October: cumulative gain of +126.57% (the highest of the year), usually the strong rebound after the bottom is completed at the end of September.

Based on this set of statistics, Chengfeng Linghang recommends “December 1” as the best annual rebalancing execution date, for these reasons:

• Wait out November’s rising season: let the upward momentum run its course first, then execute the profit-locking action, and you can lock in more of the gain.
• Avoid September’s pit: if you force rebalancing in September, you often sell near the annual low, and performance gets a chunk eaten out of it.
• Cross-year execution: December 1 is near year-end, so it can cover the complete gain/loss of the year before making a judgment, avoiding “acting too early, getting it backward, and missing the later part of the move.”

Of course, this is only a “probability reference” based on historical statistics, not a hard rule that must be obeyed. Investors may also choose another meaningful fixed date, such as: the year’s lucky day / birthday / wedding anniversary (easy to remember each year and not miss), or the anniversary of completing the asset allocation or of starting retirement. The most crucial thing is not “which day,” but “fixing it down and executing on the same day every year” — this is the most basic requirement for elevating rebalancing from subjective emotion to mechanized discipline.

And in another extreme scenario, when the market enters a frenzied “melt-up” phase — the central bank showering money to the point that funds have nowhere to go, and the stock market shoots vertically at a ninety-degree angle — the portfolio proportions will undergo violent drift. An original 433 allocation may, in a very short time, become a 352 state or an even more out-of-control proportion because of the surge in the leveraged position. At this moment, the retail investor’s greed makes them feel: “leverage is earning so fast, whatever you do, do not sell!” But the capitalist’s discipline forcibly executes the most counter-human-nature operation: sell the asset that looks the most profitable and the strongest on paper (leverage) to buy the one that looks lifeless, cash or short-term bonds.

You must know that behind a melt-up often lies the drying up of liquidity and the bursting of the bubble. Selling the wildly profitable leverage at the highest point is not because you are bearish on the future, but because you must forcibly convert this paper wealth into the real ammunition you will use to “scavenge the bodies” and save your life in the next great crash.

The Dynamic Rebalancing and New-Capital Stripping for the Employed

Although the rules of rebalancing are rigorous, for the “employed” who have continuous salary income, operation enjoys much more flexibility. Because the employed do not rely on the portfolio to produce living expenses, and possess the ability to keep injecting new capital, they do not necessarily need to forcibly execute traditional smart rebalancing every year.

If, after you have set it all up on autopilot, you do not increase your borrowing amount, then you actually do not need to bother with it; you just leave it on autopilot like that. (Video 00512)

The employed can use the monthly salary funds they invest to carry out “dynamic rebalancing.” The smartest approach is to “link rebalancing to the year-end bonus”: each year when the bonus comes, take the opportunity to review the proportions on the books, locking in profit in up years, and in down years using the bonus to add on with the trend. This approach not only saves the friction cost generated by selling assets, but also naturally pulls the portfolio back to the expected allocation proportions.

But here a technical problem arises: at the year-end settlement, how do you know which part is the gain the market grew out and which part is the new principal you threw in? If you mistake the newly invested principal for a gain and pull back 30% of it, that is equivalent to swapping from your left hand to your right, all effort for nothing! The solution is simple: when performing the smart-rebalancing calculation at year-end, you must forcibly strip “the total new principal invested during the year” out of the total market value. Only when

Year-end total market value − year-start total market value − new principal invested during the year > 0

is this the true net capital gain that arose from the market rising. You can only execute the 30% profit-locking pump against this portion of pure profit. The principal is the hen that lays the eggs, and it absolutely must not be killed by mistake in the course of rebalancing.

A worked example makes this clearer. Suppose at the start of this year your 00670L leveraged position had a market value of NT$1 million, and at the year-end settlement the books turned into NT$1.35 million — on the surface it looks like “an increase of NT$350,000.” But within that NT$350,000, you added an extra NT$50,000 of new principal midway through this year. The correct profit-locking action is: first subtract that NT$50,000 of new principal from the NT$350,000, leaving NT$300,000 as the “true net capital gain”; then execute the 30% profit-lock against this NT$300,000, that is, sell NT$90,000 and switch it into 00865B as defensive replenishment. If you forget to strip out the NT$50,000 of new principal and directly execute 30% against the NT$350,000 = NT$105,000, you would sell NT$15,000 too much, which is equivalent to pulling back a portion of the new principal you invested this year — this is the classic “left hand swapping to right hand” that kills principal by mistake.

The Down-Side Defense Mechanism and the ICU First Aid of an Extreme Bear Market

In a down year, you draw 2% from the cash position to replenish positions, and the way the denominator of this 2% is calculated directly determines the survival rate. For the “employed” who have a stable salary, the cash pool is dynamically replenished, so they can use “2% of current total assets” for the calculation, buying more shares while it is cheap. But for “retirees” who lack off-market cash flow, cash is spent one part fewer with each use, so they must adopt “2% of the original initial principal” as an absolute ceiling. This is a defensive rebalancing, ensuring that in the extreme winter of consecutive big market drops, the life-saving cash is not exhausted prematurely because the calculation base fluctuates. The supreme guideline for the retiree is to stay alive, not to bottom-fish.

Teacher James once gave, in a video, a more blunt binary standard for judging the replenishment base — “whether or not you carry pledge-borrowing”:

If you have securities-pledged borrowing, our simulation uses your assets at the start, not the assets you have grown. For example, say you started with NT$10 million, then the market rose and rose to NT$16 million, and from NT$16 million it fell down to NT$14 million. Do you use 2% of NT$10 million, or 2% of NT$16 million? Our simulation is NT$10 million, 2% of NT$10 million, not NT$16 million. So that is your original assets, for those who pledge stocks. For those who do not pledge stocks, it does not matter, you can use the year-start or the year’s highest assets too. (Video 00506)

This rule is highly consistent with the “employed vs. retiree” logic — because retirees mostly rely on pledge-borrowing to generate cash flow, the two reading standards converge on the same answer: anyone carrying pledge debt must adopt the strictest “2% of original principal” rule, locking the denominator down at NT$10 million rather than the NT$16 million after the rise.

The gap in the practical calculation is very concrete: 2% of NT$10 million is NT$200,000 of replenishment, while 2% of NT$16 million is NT$320,000 — for a pledge-carrying retiree who lacks off-market cash flow, this NT$120,000 gap would rapidly exhaust the life-saving cash across consecutive bear years. An employed investor without pledging, because cash flow keeps being injected and there is no threat of forced liquidation, can adopt the more aggressive “2% of current assets” method to increase the down-year replenishment force.

▲ Figure 12-4 The down-year catch mechanism, where cash is injected into the leverage container

However, in practical operation, what the retail investor most often faces is not an epic-scale stock-market crash, but a minor market pullback (for example, a mere 5% drop). At this moment, discipline tells the investor that the 10% add-on line has not yet been reached, but inwardly they intensely want to act, itching so badly they cannot sleep. Do not fight human nature by forcibly holding back; the capitalist knows how to use a mathematical mechanism to “scratch the itch.” At this point, you can execute a small-scale “leverage conversion”: sell NT$2,000 of the underlying broad-market ETF and turn around to buy NT$1,100 of the 2× leveraged ETF.

Calm down and do the math: the market’s actual exposure rises slightly from NT$2,000 to NT$2,200, secretly satisfying the desire to “add on the dip”; and even more marvelously, out of thin air you now have an extra NT$900 of cash defensive level in your pocket! Using a tiny structural conversion in place of tapping the off-market life-saving reserve both soothes the brain’s panic and holds firm to the bottom line of fund control; this is a high-level operation for mastering human nature. (This is a non-standard itch-scratching technique, not a standard annual rebalancing; if the underlying is in a taxable account, selling may generate a tax liability.)

In an extreme bear market, if the cash reserve is less than 15 years of living expenses, the traditional withdrawing of cash to replenish positions becomes extremely dangerous, and could exhaust the oxygen before dawn. At this point you can activate an advanced extreme operation: “rebalancing without depleting cash.” The concrete method is not to touch cash, but to convert the underlying (QQQ) directly into 2× leverage (QLD). This can increase market exposure by raising the “leverage content” of the holdings, without consuming the margin of safety. The cash position still firmly guards the survival bottom line, but the attacking force has switched to stronger equipment; this is the cold-blooded discipline of “able to attack yet not die in retreat” in a desperate situation. (This is a non-mainline extreme emergency measure, limited to the critical scenario of insufficient cash, and is not a routine exception to the core position of “hold the underlying and never sell”; when the underlying is in a taxable account, the conversion may generate a tax liability, and ordinary annual rebalancing still centers on the standard operation between cash/short-term bonds and leverage.)

Flexible Rebalancing, an Advanced Variant from the 00662 Community (Not Mainline)

On top of CLEC’s core smart rebalancing, PonPon, a member of the 00662 community, inspired by a passage from Teacher James, developed an advanced variant, “flexible rebalancing.” It takes “whether the cash-like position has accumulated a full 15 years of living expenses” as the watershed: when the cash cushion is not yet full, the down-year replenishment switches to selling the underlying (QQQ) rather than tapping cash to top up leverage, guarding the life-saving cash; when the cash cushion is full, the up-year profit-locking switches to shifting into the underlying rather than into cash, letting the total-asset Beta descend gently and retaining more bull-market participation momentum. This variant is then subdivided by the destination of the locked profit into a “conservative type” and an “aggressive type.”

Note that flexible rebalancing is PonPon’s personal conception, its backtest covers only a single historical interval (2000 to 2025), and the original author states outright that it is not yet a settled conclusion, so this book does not list it as mainline. The complete three-type operational comparison, design logic, and source are compiled in “Appendix 7, Advanced Allocation and Rebalancing Tactics.”

The Precision Method for Investing a Large Lump Sum

When a large amount of funds that will shake the overall asset proportions (for example, applying for a NT$3 million personal loan, or receiving a large year-end bonus) comes in, you must not buy in batches by feel, but must use the precision method to pull the portfolio back to the ideal allocation proportions in one shot. (Chengfeng Linghang, “EP04”)

• New target asset amount

(Current total market value + total new funds invested) × the target proportion of that item

• Actual amount to buy

New target asset amount − the current market value of that item

Taking the 442 allocation (underlying 40%, 2× 40%, cash 20%) as an example, suppose the current total market value is NT$33.45 million and you are about to invest a large lump sum of NT$3 million:

• New total-asset base: NT$33.45 million + NT$3 million = NT$36.45 million
• Underlying target to reach: NT$36.45 million × 40% = NT$14.58 million; if the current underlying market value is NT$10.8 million, you need to buy NT$3.78 million.
• 2× target to reach: NT$36.45 million × 40% = NT$14.58 million; if the current 2× market value is NT$11.1 million, you need to buy NT$3.48 million.
• Cash target: NT$36.45 million × 20% = NT$7.29 million; if the current cash is NT$11.55 million (that is, 3,345 − 1,080 − 1,110), then you should draw NT$4.26 million out of the existing excess cash (1,155 − 729), which together with the NT$3 million of new funds totals NT$7.26 million, used to simultaneously buy NT$3.78 million of underlying and NT$3.48 million of 2×, correcting the proportions in one shot.

Through this formula, no matter how the asset proportions have drifted in the past, as long as the deployable funds (new funds + excess cash position) are large enough, a single investment can instantly correct the portfolio back to the most perfect allocation proportions, without selling any existing holdings.

The Autopilot System That Lets the Brain Rest

The greatest enemy of investing is always the investor’s own greed and fear. On a capital racetrack full of the unknown and of risk, no one can forever precisely predict the weather and road conditions. Once you have set the proper asset proportions (for example, the golden allocation of “underlying : leverage : cash/short-term bonds” at 4:3:3) and have made clear the operating rules of rebalancing, the remaining task is to activate “autopilot.”

After you have finished setting your asset allocation and start borrowing to use it, once you do smart rebalancing, that is autopilot. The proportion relative to your assets is no longer that important now, because we have simulated it so that your debt maintenance ratio stays at 167% or above. (Video 00498)

This system is like a strict disciplinary contract you sign with yourself. Each year you need only spend a very small amount of time, mechanically executing the actions according to the formula; the rest of the time, your life ought to be completely emptied out, to experience life, focus on your main career, or accompany your family. By drawing a clear strategic red line between the core underlying assets and the dynamic components (leverage / cash), the investor has effectively installed the toughest shock absorber for their own wealth. Guarding a 15-year cash level and strictly obeying the smart-rebalancing formula is not only the economic moat that protects assets from going to zero, but also the sole remedy on the road to peace of mind and financial freedom.

In practice, just how low is the time cost of this “autopilot”? The answer is: one minute a year. Fix on a point in time you will not forget (for example, the day you receive your mid-year bonus, your birthday, or New Year’s Day each year), spend one minute logging into your account to confirm the current market value of each position, use the formula to calculate how much to sell / how much to buy, place the orders, and the whole year’s rebalancing is done. For the remaining 364 days, 23 hours, and 59 minutes, you do not need to watch the market at all, do not need to chase the news, and do not need to be anxious over the market’s highs and lows. “Managing your finances takes only one minute a year” is not an exaggerated slogan, but the true state after you have asset allocation + smart-rebalancing discipline. Fill up the time for learning, and reduce the time spent watching the market to zero — this is the highest realm of giving life back to living.