The Beginner’s Threshold and the 433 Allocation

Risk disclosure: The various allocation formations described in this chapter (433 / 442 / 703 / 505 / 514 / 424 / 226 / 3313, and so on), the drawdown rates (2% / 4% / 7%), the cash multiples, and the Beta ranges are all backtested compilations of publicly available historical data and academic explorations of asset allocation. The conclusions of a historical backtest do not mean the future will necessarily repeat them; an individual investor’s optimal allocation depends on factors such as age, asset size, cash flow, and risk tolerance. Nothing in this book constitutes individual investment advice, and readers should make their own independent judgments about their actual allocation decisions.

The core of asset allocation is not how much you earn when markets rise, but how you survive an extreme crash. The core of managing money is not to pursue maximum returns, but to make sure you are never wiped out by any single extreme event. So the order of thinking must be reversed — not first asking how high the return can go, but first asking how much risk you can bear: first set the cash proportion and draw the risk boundary, then talk about Beta and the possibility of amplifying returns. Before entering any allocation, a cash reserve (an emergency fund) must be in place: a beginner should have at least 6 to 12 months of living expenses ready; a retiree needs to keep a higher cash / short-bond defensive line, scaled to asset multiples and the drawdown rate (as in the later example of NT$50 million / 50× annual expenses, where 30% cash roughly equals 15 years of annual expenses — though not every retiree needs a fixed 15 years).

A word especially for young people who are just starting out: if your total assets are still below 3× your annual expenses, you cannot directly apply the 30%-cash proportional rule. For example, a young person whose annual expenses are US$100,000 but whose total assets are only US$160,000 must first lock the full US$100,000 down as a “one-year absolute emergency fund,” leaving only the remaining US$60,000 available to invest. Only once assets grow past US$500,000 (30% × US$500,000 = US$150,000, which already exceeds the one-year emergency-fund floor) does one formally transition to the regular proportional allocation of “30% cash.” Survive first, then scale up — this is the most important discipline for a beginner.

Just like Han Xin marshaling his troops, the more the better, and it must be centrally managed. Many beginners are used to scattering their capital across various odd accounts, or trapping their capital in individual stocks, low-return endowment insurance, and cash-flowless property. This stagnant water, which cannot generate efficient compounding, should be inventoried item by item and, after deducting taxes, surrender costs, and necessary protection, gradually liquidated and moved into the high-quality index assets that fit this book’s core logic. Counting up all your capital and concentrating your firepower is the first step in starting up the engine of capital operation.

You must count all of your capital clearly and consolidate it into a single account. (Video 00507)

Beyond consolidation, in practice you can also set up a “twelve-month laddered time-deposit method”: split the first year’s repayment amount into 12 tranches of time deposits with maturities of 1 to 12 months, so that the repayment cash flow is completely decoupled from stock-market volatility. This survival money is your oxygen during a bear market.

An investor can treat 00662’s long-term annualized return of 12% as a “benchmark mirror for long-term performance.” When a financial adviser pushes endowment insurance or a high-dividend fund, simply take the final payout amount they claim and back out its annualized return. If the number you get cannot even see the taillights of 12%, stand up and walk away at once. The reason the NASDAQ-100 is so formidable is that the index rules deliberately exclude the financial sector, letting growth momentum concentrate on innovative enterprises rather than the cycle of bank interest-rate spreads; across different rate environments it retains a purer growth inertia than the S&P 500. But this does not mean it is immune to interest rates — growth assets are still pulled along by the discount rate.

The Three Elements of an Asset Portfolio

Life needs the three elements of sunlight, air, and water to sustain itself; asset allocation likewise needs three elements — the underlying, the leverage, and the cash — none of which can be missing. Using a fund that tracks the NASDAQ-100 index as the underlying core not only avoids the risk of a single stock going to zero, but also, through the index’s mechanism of weeding out the weak and keeping the strong, forever holds the collection of the strongest enterprises.

Through the “433 allocation” (40% underlying, 30% leverage, 30% short-bond), the overall Beta is 1.0 (0.4 × 1 + 0.3 × 2 + 0.3 × 0). The magic of this allocation is that using only 70% of your capital, you achieve market exposure roughly equal to being 100% fully invested in the broad market (the Beta is approximate, not identical in return and fees), while still holding a steady 30% cash ammunition in hand. When a big drop hits, you can execute “smart rebalancing,” moving cash into the leveraged position, achieving a true stance of being able to advance and attack or retreat and defend.

The Allocation Philosophy of Polarized Offense and Defense

Before setting specific proportions, an investor must understand the philosophy behind the 433 allocation — the allocation logic of “polarize the two ends, leave the middle a vacuum.” The core of this line of thought is to thoroughly discard the mediocrity of the SPY “mean reversion” approach, refusing to let capital waste away in inefficient, middle-of-the-road assets. Capital is placed only at two extremes: at one end, “extreme offense (QQQ/QLD),” and at the other, “extreme defense (a high proportion of cash / 00865B).” Those mediocre assets in the middle that are neither strong enough nor stable enough (such as high-dividend products, long-term bonds, and conservative funds) are not touched at all. Never use mediocre assets to lower the ceiling of the portfolio; instead, control the overall Beta and drawdown risk through dynamic adjustment of the cash proportion, ensuring the strongest penetrating power in a bull market and the thickest cushion in a bear market. This “black or white” allocation is the only path to breaking mediocrity and crossing the class divide.

This line of thought of “placing capital only at the two ends and discarding the middle” is shaped exactly like a “dumbbell” — heavy at both ends, hollow in the middle, with capital concentrated at the two extremes of maximum offense and maximum defense, using the cash proportion to control overall volatility rather than using mediocre assets to hold down the profit ceiling. Many retail investors are superstitious about total-market indices, believing that holding thousands of stocks is what safety means, but quantitative research shows that once holdings increase past a certain number, the marginal benefit of diversifying away non-systematic risk falls rapidly, and expanding further mainly just moves you closer to the market average, unable to eliminate systematic risk any further; to save a little on capital-gains tax by clinging to mediocre assets, what you pay is a far more expensive “time tax.”

The endgame goal of this polarized allocation can be examined against a “four-type” framework. In traditional investing, risk and return roughly show a symmetric relationship, which can be summarized into four typical states:

• Type one is “rising and falling in lockstep with the broad market,” simply holding QQQ over the long term without making any further allocation — the most basic index accumulation.
• Type two is the add-Beta route of “falls more, rises more,” holding a large amount of leveraged ETFs such as QLD / TQQQ, sacrificing downside tolerance in exchange for upside explosiveness.
• Type three is the reduce-Beta route of “falls less, rises less,” with cash or short bonds making up as much as 50% or more of the portfolio; although it resists drops, it also shaves off the growth momentum of bull-market years.
• Type four is the ideal end state that is pursued — “falls less but rises fast,” using a cash line of defense to withstand extreme crashes, then combining the explosiveness of leveraged funds with the smart-rebalancing discipline of “add leverage in down years, lock in profit in up years,” forcibly manufacturing asymmetry out of thin air.

Type four is not a gift the market gives you, but a result “engineered” through discipline and mathematics. When constructing your own allocation, you can use this four-type framework to check your own work: if your portfolio rises and falls in lockstep year after year (type one), it means you have not yet entered a polarized allocation; if it soars and crashes wildly (type two) or is so conservative that it misses the bull market (type three), it means the Beta setting is wrong; only when you truly enter the dynamic state of type four — “a thick downside line of defense plus fine-tuned upside leverage” — do the two ends of offense and defense truly mesh and operate together.

The Historical Evolution of the 70/30 Allocation

Almost all of the current allocations in the CLEC system (703 / 433 / 442 / 505, and so on) can, at their underlying proportion, be traced back to the original model of “70% stocks + 30% cash.” Understanding its evolutionary path makes clear why “30% cash” is a hard floor that appears again and again.

The first stage is the plain-and-simple state of “no pledging, selling stock to obtain living expenses.” Take annual expenses of NT$1 million and retirement assets of NT$50 million as an example: NT$35 million in stocks + NT$15 million in cash (15 years of annual expenses), living off stock sales and using the cash position to avoid being forced into fire sales of stock during a bear market. This is the origin of the original 70:30 proportion.

The second stage introduces the tool of “stock pledging.” Teacher James found that a pure stock-selling strategy has a blind spot — the US has a capital-gains tax, selling stock immediately triggers the tax, and the sold shares permanently exit the compounding process. After switching to obtaining living funds by pledging stock as collateral (securities-backed loan / pledging), because the core assets are not sold, this usually does not immediately trigger the capital-gains tax on stock sales, and the principal can keep growing (but this is not entirely tax-free — you still have to account for dividend tax, interest cost, estate tax, and so on). The 70:30 structure stays unchanged, but the source of living expenses changes from “selling stock” to “borrowing via pledge.”

The third stage introduces “leveraged stock.” Since stock can be pledged to avoid the erosion of selling, can the cash position also be made to serve a greater value? At this point the concept of 2× leverage is introduced (50% leverage + 50% cash, whose investment return equals that of 100% underlying while still retaining half in cash), upgrading the pure “cash cushion” into an equivalent underlying of “leverage + cash.” This is the mathematical basis of leverage-containing formations such as 433 / 442.

Looking back at this history reveals one important discipline: you must have a pledging system in place before you can add leverage. If you do not possess pledging (to avoid being forced to sell stock during a big drop), you can only stay in the first-stage, plain-and-simple version of 30% cash (that is, the 703 or 802 allocation). This causal relationship cannot be reversed — any practice of “adding leverage without pledging” skips the second stage and forcibly connects it to the third, equivalent to dismantling the entire line of defense in exchange for explosiveness, and it will be knocked back to square one at the first bear market.

Having discussed the skeleton of the allocation, Teacher James, in Chapter 3 of his 2026 “Billion-Dollar Investment Lecture,” wraps it up in one sentence — do not make the allocation too complicated; the real point is to “keep enough cash, and buy in immediately when you have money”:

The most important thing in asset allocation is to stay alive. Cash is air, so if you keep hesitating over whether to buy in, you might as well keep enough cash — the cash position you feel comfortable with — and for everything else, just buy in at market price immediately. So asset allocation is really just a garnish; the point is to be roughly and vaguely correct. Once the asset allocation is done, get down to it quickly. (Video 00696)

In other words, the earlier “polarization and four-type framework” is meant to help you understand the skeleton of the allocation, but when actually executing, you do not need to chase a perfect proportion — as long as your cash level is reassuring enough and the overall Beta lands in a range you can bear, you should decisively make a lump-sum entry and leave the rest to time. Excessively agonizing over “should I wait a bit more, is the proportion off by a few percentage points” instead leaves capital idle and misses out on compounding. This also echoes the book’s core discipline that a “vague correctness” beats a “precise error.”

Setting the 442 and 433 Golden Allocations

After completing the consolidation of capital, we next clarify the two main allocation logics. The setting of the cash proportion is closely tied to the investor’s career situation and cash-flow needs.

Before selecting a proportion, first confirm “who this allocation is for” — the CLEC allocation is a ladder from conservative to aggressive; different people have different starting points:

• The general public (most people): start with 80/20 or 70/30, pure underlying plus cash, no leverage, putting “staying alive, not being washed out” first.
• Those still accumulating assets: advance to 433 (40% underlying, 30% 2× leverage, 30% cash), using limited leverage in exchange for accumulation speed.
• The aggressive life-cycle investing approach (under 30 years old): only then consider 442 (40% underlying, 40% 2× leverage, 20% cash), this most aggressive formation; for those over 30, 433 is already the most aggressive reasonable ceiling, and adding on to 442 is not recommended.

The further the numbers move toward “442,” the younger the required age, the steadier the cash flow, and the greater the volatility one must be able to bear; the correct starting point for the overwhelming majority of readers is 70/30 or 80/20, not chasing the high slope of 433 or 442 right out of the gate.

442 is for someone who does not need to take money out of their stock support to use 442. Someone who has to take money out of their assets, who does not have a job producing stock, can only use 433. (Video 00502)

For younger investors still fighting in the workplace, who have stable salary income and do not need to rely on their portfolio to pay living expenses, the 442 strategy (40% underlying, 40% 2× leverage, 20% cash) offers a route to aggressively accumulate assets. The Beta of this allocation is about 1.2, and because the cash proportion is lower and the leverage proportion higher, its volatility in the face of market shocks is extreme. For this stage, the most robust strategy is to “open the pledge line in advance but not borrow.” This ensures that while enjoying the full compounding of the market during the accumulation period, you still hold in hand a cash-flow protective umbrella that can be opened at any time, to cope with sudden unemployment or a market crash. By Teacher James’s line, 442 is an aggressive choice within the life-cycle investing approach, better suited to young people under 30 who are still decades from retirement; for those over 30, even those employed with a salary, 433 is recommended as the aggressive ceiling.

Before getting into specific proportions, we must first warn of a human-nature trap that is easily overlooked — the “1/N law.” The “naive diversification” research of behavioral economists Richard Thaler (2017 Nobel laureate in economics) and Shlomo Benartzi found a peculiar phenomenon: as long as an investor faces two or more choices simultaneously and can select more than one, the subconscious tends to distribute capital “evenly” across each option, rather than making a reasonable allocation based on the risk characteristics of the assets. One piece of empirical data from a University of California employee retirement plan revealed the power of this effect: when employees faced a mix of 1 stock-market fund and 4 bond funds, the average stock-market allocation was only 34%; but when the mix was reversed to 4 stock-market funds and 1 bond fund, the employees’ stock-market allocation shot up to 75%. The same group of people, facing the same question of “how much should stocks account for,” gave answers that turned out to depend on the number of each type of fund on the menu.

This blind spot is especially devastatingly costly in retirement accounts. If the 1/N instinct drops the stock-market allocation from 80% to 30%, over 20 years the retirement fund will earn 25% less; stretched out to 30 years, the gap widens to 35% to 40%.

Without going through scientific analysis and instead “splitting evenly by instinct” can cause an investor to lose nearly half of the wealth that should rightfully be theirs by the time they retire. The reason the CLEC system’s specific proportions such as 442 / 433 / 802 matter is not that the numbers “3” or “4” themselves are magical, but that these proportions have been validated by 25 years of historical backtesting plus 100,000 Monte Carlo simulations, making them the mathematically optimal solutions that “can still survive an extreme stock crash.” Any even-split instinct such as “I feel more at ease splitting into three parts” or “splitting evenly across four ETFs looks fair” is essentially an extension of the 1/N trap — it feels reasonable, but in the data it is slow suicide.

By contrast, the 433 strategy (40% underlying, 30% 2× leverage, 30% cash) is the “golden ratio” among all allocations, and also the survival floor by which most people can safely get through an extreme bear market. This allocation has a Beta of 1.0, meaning its overall volatility is comparable to holding 100% underlying funds, yet the investor holds as much as 30% cash in hand as defensive armor. This 30% cash reservoir is the sea-stabilizing needle that supports pledged borrowing and maintains retirement living cash flow.

At the start, the asset allocation’s cash must exceed 30%; even more aggressively, it cannot exceed the underlying fund, and cannot exceed the cash. (Video 00519)

The Advantages of 00865B Over 00864B (and 00859B)

There are three main choices of Taiwan-listed US short-term government bond ETFs: 00864B (CTBC US Treasury 0-1), 00865B (Cathay US Short-Term Treasury), and 00859B (Capital 0-1 Year US Treasury). What all three track is essentially “US Treasury bills of 0 to 1 year,” but because of differences in “whether or not they distribute” and in “size / trading volume,” the CLEC system, in its allocation logic, leans toward the only “non-distributing” one, 00865B (because not distributing avoids triggering the second-generation NHI supplementary premium). A comparison of the basic data of the three follows (Source: simplified prospectuses of 2024/12/31):

The differences in return and expense ratio among the three are extremely small (less than 1 percentage point), and 00864B is the largest in size and trading volume. Yet Teacher James still explicitly recommended in an April 2025 course that Taiwanese investors switch to 00865B — the main reason being precisely that 00864B has, since May 2025, implemented a quarterly distribution scheme (with distribution records already in 2025/08, 2025/11, 2026/02, and 2026/05). This change turns 00864B from an “accumulating” type into a “quarterly-distributing” one, immediately producing the same tax and compounding friction as 00859B.

The most central reason for moving capital from 00864B to 00865B is that 00864B has already begun distributing dividends, and this brings long-term investors much inconvenience and erosion in terms of taxes, transaction costs, and management. 00859B is likewise a quarterly-distributing structure with the same problem — among the three, only 00865B maintains the structure of “NAV accumulation, non-distributing,” making it the most qualified under CLEC’s priority conditions of “non-distributing, auto-accumulating, reducing distribution friction” (the actual choice still needs to consider size, trading volume, and personal taxation). The detailed reasons for recommending the switch to 00865B are as follows:

Avoiding the Second-Generation NHI and Overseas-Income Tax

The dividends distributed by 00864B are overseas income, and when the capital position is large (for example, exceeding NT$1 million), it may face overseas-income tax issues, and receiving dividends may also be subject to withholding of the second-generation NHI supplementary premium. Because 00865B does not distribute, it avoids the second-generation NHI supplementary premium triggered by distribution itself, as well as the problem of manual reinvestment; but this does not amount to being entirely tax-free — the future capital gains on sale, overseas income, or the alternative minimum tax still depend on personal circumstances.

Saving the Transaction Fees of Manual Reinvestment

To maintain the compounding effect, after receiving 00864B’s dividends you must manually buy back in to reinvest, incurring an additional transaction fee each time, and it must be done continuously every month. 00865B’s returns are directly reflected in the NAV, automatically achieving compounding with zero manual intervention.

Avoiding the Interest Erosion of Idle Capital

Every day you delay manually reinvesting a distribution is a day of interest earned less. Capital that is not rolled back in due to forgetfulness or the hassle will, over the long run, lose the power of compounding. 00865B reflects returns directly in the NAV, with no window of idle capital.

A True Shield of Defense

Many people mistakenly believe that high-dividend ETFs (such as 00713) resist drops, but during the 2022 stock crash 00713 too plunged sharply and even hit limit-down. Because 00865B has hedging properties, when the stock market drops hard it often rises against the trend as the US dollar appreciates, making it the truly strongest shield that protects the emergency fund from being eroded. Looking ahead to 2026, as the US “Genius Act” is promoted, compliant stablecoins (such as USDC) become digital dollars pegged to US short-term Treasuries; but they still carry risks such as the issuer, wallet private keys, platform collapse, redemption liquidity, and regulatory change, and are not equivalent to bank cash or a short-bond ETF. At most they can serve as a tiny-proportion research / backup tool, and should not be regarded as a risk-free safe harbor.

Many investors are drawn by the high yield of “high-distribution corporate-bond ETFs” (such as 00948B, an investment-grade corporate bond), mistakenly thinking corporate bonds can also play the role of a cash line of defense. But breaking it down at the underlying level of financial engineering, corporate bonds and US government bonds have a structural difference in defensive capability — the key lies in the “credit spread.”

Corporate-bond yield = risk-free government-bond yield + credit spread

This credit spread is the extra compensation an investor demands for bearing a company’s default risk. In a bull market, the credit spread is usually very small, and corporate bonds look much like government bonds; but when downward signals appear in the economy and the market’s fear-driven expectations of corporate failures rise, the credit spread expands sharply, and this expanding spread directly eats up, or even more than reverses, the capital gains brought by rate cuts, causing corporate-bond prices to crash in sync with stocks during a market crash, completely losing their hedging function.

The 2008 financial tsunami, the 2020 COVID crash, and the 2022 rate-hike bear market all verified the same conclusion: at the very moment you need defense the most, corporate bonds will betray you; only pure risk-free government bonds (such as BOXX, SGOV, 00865B) can, under the double pressure of a stock crash and an expanding credit spread, still hold their NAV or even rise against the trend. This is the financial-engineering basis for why the CLEC system strictly stipulates that the cash line of defense can only use US short-term government bond products, and absolutely excludes any “high-yield corporate bonds,” “investment-grade corporate bonds,” or “non-investment-grade bonds (junk bonds).”

An investor should regard the paper loss caused by exchange rates as a “necessary cost” and ignore short-term exchange-rate fluctuations. A large appreciation of the NT dollar is usually caused by foreign capital pouring in and buying heavily; at such times equity assets (00662 or 00670L) will inevitably be accompanied by a surge, and the few percent of paper exchange loss on 00865B will be completely covered by the gains on the stocks. Conversely, during a stock crash the NT dollar depreciates, and short bonds then perform their exchange-rate-shield function.

The most common mistake is to want to swap 00865B out just because it shows a paper exchange loss. But as long as the stock assets grow by NT$600,000 over the same period, even if 00865B erodes by tens of thousands, total assets still grow substantially. Do not give up the structural stability of the entire allocation for the sake of one localized paper loss.

Why Allocate to 00865B Rather Than Cash or a Time Deposit

In the CLEC allocation, 00865B is the “golden supporting actor” of asset allocation. Readers often ask: “Since it is for hedging or parking temporary capital, why not just use a bank time deposit?” Three core differences explain why 00865B cannot be replaced by cash / a time deposit:

• Strategic difference: the auto-pilot of rebalancing. Cash / time deposits are separated from the securities account — when the stock market drops hard and you need to manually transfer money from the bank or break a time deposit, there is enormous psychological resistance (human nature does not like putting money in during a market panic); whereas 00865B is right there in the securities inventory, and when the drop is deep the market-value proportion of 00865B automatically rises, so you only need to disciplinedly “sell 00865B, buy 00662” to complete a buy-low-sell-high, an operation that is smooth and overcomes human nature.
• Return difference: cash-like assets that fight inflation. Time-deposit rates are, over the long run, below inflation; 00865B tracks US short-term Treasury bills, with an annualized return of about 2% to 3% or more (depending on the FED rate), and although there is an extremely small degree of fluctuation, it can earn a better “holding return” than a time deposit while the capital is resting.
• Functional difference: asset activation and risk control. 00865B is an excellent collateral for pledging — its price volatility is extremely small, and when a stock crash causes 00662’s maintenance ratio to drop, 00865B can stabilize the maintenance ratio of the entire account and prevent a forced liquidation. In terms of taxation, 00865B is overseas income (the bond interest rolls into the NAV), whereas time-deposit interest above NT$20,000 is subject to withholding of the second-generation NHI supplementary premium.

For the small-capital crowd (who already hold 00662, 00670L, and so on), the core reason to put temporary capital into 00865B is to “build discipline plus protect the position” — keeping capital on the field, ready at any time to execute rebalancing when the stock market pulls back; the 2× leveraged ticker is highly volatile, and pairing it with 20% of 00865B can greatly reduce overall psychological pressure. Cash or a time deposit is a safe asset, but 00865B is an asset that “breathes together with the investment strategy,” making wealth accumulation more systematic and efficient.

The 613 Starting Path for the Small-Capital Crowd in the Accumulation Period

Those who have just stepped into the market, with total assets not yet accumulated to NT$1 million, often wonder “should I go straight to 433 from the start.” In practice, a more robust path is to “start with 613 first, accumulate to NT$1 million, then switch to 433.” There are three concrete steps:

• First step: first lock down 6 to 12 months of living expenses as an emergency fund; this money does not enter the market, does not participate in any allocation, and serves as a pure line of defense against unemployment or a major expense;

• Second step: continuously invest your investable capital on a regular, variable-amount basis at the 613 proportion (60% underlying + 10% leverage + 20% cash + 10% reserved rebalancing ammunition); at this stage, continuously accumulating by proportion is itself a form of dynamic rebalancing — new cash flow is naturally directed toward the positions that have deviated from the proportion, without needing to deliberately make extra selling moves.

• Third step: once total assets accumulate past NT$1 million (at which point 30% cash is already enough for the one-year emergency-fund threshold), you can choose to keep maintaining 613, or formally upgrade to 433 and begin annual rebalancing. Personal-loan repayment must still come from work salary — this discipline does not change from the very first day of the accumulation period until before retirement.

▲ Figure 11-1 The three-step path for the small-capital crowd in the accumulation period — lock down the emergency fund, start with 613, switch to 433 at NT$1 million

Asset Allocation Formations at Each Tier and Their Applicable Groups

For different risk tolerances and life stages, asset allocation must display extremely high flexibility. The table below defines four core defensive formations in the CLEC investing system:

For beginners just stepping into the market or the retirement crowd, the “703 allocation (Beta 0.7)” is regarded as the entry-level first choice. If pursuing higher efficiency in the accumulation period, you can adopt a “Beta 1.1 precision model,” breaking the traditional myth that “keeping cash must lower returns”:

A common blind spot for beginners: when you have already adopted the 703 allocation and later want to add a leveraged ETF (such as QLD / 00670L) to become 613 or 433, do you have to wait until the market drops before executing? The answer is “no need to wait.” “Changing the overall allocation strategy” and “the annual rebalancing discipline” are two completely different things. The former is a system upgrade, which can be adjusted directly at any point in time (for example, turning a 10% cash position into a 10% leverage position upgrades 703 to 613); the latter is an annual action, which is the only thing that needs to follow the discipline of “add leverage in down years, lock in profit in up years.” Clarifying this distinction lets a beginner, when “wanting to upgrade leverage,” avoid getting stuck in the market-timing myth of “should I wait for a drop.” The CLEC system never times the market; whether it is the initial allocation, a system upgrade, or annual rebalancing, none of them require predicting highs and lows.

70% QQQ + 20% QLD + 10% cash → Beta = (0.7 + 0.4 + 0) = 1.1

If higher defense is needed, upgrade tier by tier according to the drawdown rate:

424 allocation: 40% underlying, 20% 2×, 40% cash → ensures cash is more than 2× total liabilities

514 allocation (Beta ≈ 0.7): 50% underlying, 10% 2×, 40% cash → suits retirees with a drawdown rate of 3% to 4%, retaining a growth spark while maximizing the cash line of defense

415 allocation: 40% underlying, 10% 2×, 50% cash → suits a drawdown rate of 4% to 5%, with asset safety as the supreme guideline, sacrificing explosiveness in exchange for the absolute right to survive

For elderly seniors, or those whose asset level has already reached extreme safety, the “505 extreme retirement allocation (Beta 0.5)” offers the ultimate defense.

You retirees should use 80/20, or 70/30 is fine too; if you do not pledge, you cannot use a leveraged fund. (Video 00512)

The 345 Cash Rule and the No-Leverage Allocation

Here we can summarize a “345 cash rule”:

• 3% drawdown rate: allocate 30% cash (corresponds to the 433 formation)
• 4% drawdown rate: allocate 40% cash (corresponds to the 514 formation)
• 5% drawdown rate: allocate 50% cash (corresponds to the 505 formation)

The 345 rule above is built on the basis of “the portfolio containing 2× leverage” (433 / 514 / 505 each carry a 30% / 10% / 0% 2× position). For conservative retirees who use no leveraged funds at all and hold only the two assets of “underlying ETF + cash-like” (for example, seniors over 70, or investors whose psychological makeup is not up to leverage), the CLEC system additionally provides a corresponding no-leverage pledging allocation:

If your asset allocation is 80% underlying, that is 00662 or QQQ, and 20% money-market-fund cash-like, then each time you just directly borrow 2% and do a brainless rebalance once a year, this can pass. The second is, if your drawdown rate reaches 3%, then you can do 65% underlying fund and 35% cash. If your drawdown rate reaches 4%, then your underlying is 50% and your cash is 50%. For those with a drawdown rate above 4% — that is, assets below 25× annual expenses — it is recommended you start planning to use QQQI to replace the cash flow, and stop pledging. (Video 00685)

Compiled into a comparison table:

Readers can choose for themselves according to “whether or not to use leveraged funds” — with leverage, take the 433 / 514 / 505 main line; without leverage, take the 8/2, 65/35, 5/5 two-asset allocation. The two routes stand side by side within the CLEC system.

For those with large short-term rigid spending needs, such as the payment demands of a pre-sale (off-plan) apartment, the “226 allocation” is recommended: 20% underlying, 20% 2×, 60% short-bond. The high 60% short-bond position is a strongbox for construction-stage payments, close in nature to US-dollar short bonds (still carrying exchange-rate, interest-rate, and premium/discount risk, not a bank time deposit), which will not shrink substantially as the stock market drops, ensuring that while enjoying the market’s dividends, you will never be forced into a fire sale of assets because of an urgent need for money.

Reinforcement Formations for Those with Special Cash-Flow Needs

For retirees who need stable cash flow but still hope to maintain leverage momentum, you can refer to the “3313 allocation” variant — 30% QQQ, 30% QLD, 10% QQQI, 30% cash. This is a safety net designed for those with insufficient cash flow who nonetheless need to pledge.

QQQI can self-produce about 1% of cash flow (calculated at a conservative 10% yield), halving the growth rate of the pledged debt. Note that this is not an upgrade of 433, but a risk-control downgrade scheme. Take total assets of NT$10 million as an example: originally the retiree needs to borrow NT$200,000 a year via pledge (a 2% drawdown) to cover living costs; in the 3313 allocation, NT$1 million of QQQI can generate about NT$100,000 of distributions a year (a 10% yield), cutting the pledging need from 2% in half to 1%. Over 30 years, the accumulated pledged debt is halved, the maintenance-ratio line of defense is thicker, and the probability of forced liquidation drops markedly.

Structurally, 3313 is simply taking the standard 433’s underlying QQQ from 40% down to 10% and putting the difference into QQQI, with leverage and cash completely unchanged:

Since QQQI’s Beta is about 1.0, close to QQQ’s, this swap barely changes the portfolio’s overall volatility — what it lowers is not the stock-price volatility, but the pace at which pledged debt accumulates: relying on that 10% of QQQI’s self-produced distributions to borrow less via pledge, the drawdown need is pressed down, and the maintenance ratio is naturally thicker. The cost is sacrificing 10% of QQQ’s long-term growth in exchange for the safety margin of “borrowing less via pledge,” so it only suits people who “still want to pledge but want to relieve the pressure.”

If you are unwilling to even use pledging, and your assets are even tighter, 3313 does not apply, and you should switch to the rescue allocation below in which QQQI carries the lead. The two do not conflict but are different landing points on the same “pledging willingness × asset sufficiency” ladder, with the key to the split being “whether you still pledge” and “the QQQI proportion”:

For those whose assets are insufficient and whose estimated cash flow cannot cover retirement living, the CLEC system offers the following “QQQI rescue allocation”:

Reminder: For those whose assets are below 25× annual expenses, you can allocate 10× annual expenses to QQQI (using its conservative roughly 10% high distribution to cover living costs), with the remaining 5× annual expenses allocated to 00662 and 00865B to maintain growth. This allocation is limited to retirement emergency use, and is strictly forbidden for young people, otherwise it will severely sacrifice long-term compounding growth.

If we reverse-calculate more intuitively by “drawdown rate”: suppose total assets of NT$1 million and an annual drawdown of 7% (NT$70,000 a year), then about NT$700,000 needs to go into QQQI (7% ÷ 10% = 70%) for the distributions to fully cover living expenses; the remaining NT$300,000 is then allocated between QQQ and cash-like at a 70/30 proportion, maintaining long-term asset appreciation. The same formula also applies to a 4% or 5% drawdown rate, and readers can plug in their own actual drawdown proportion to calculate.

Twenty Thousand Monte Carlo Simulations Validate 433

Many investors, out of fear of leverage, will instinctively think that “lowering the leverage proportion” can buy higher safety. For example, adjusting the allocation to the 613 strategy (60% underlying, 10% leverage, 30% cash). Yet scientific data gives an answer that completely overturns intuition.

According to the research findings of Chengfeng Linghang’s “X000 — Monte Carlo Simulation Report V1.1.0,” in a high-intensity stress test conducted on US-stock historical data from 1999 to 2025, the system randomly generated twenty thousand Monte Carlo simulation paths, each 26 years long, with tickers spanning the underlying, leverage, and cash / short-term US Treasuries (BOXX / SGOV / 00865B), to test the true performance of various allocations under both extreme and normal markets.

This simulation system adopts three rigorous underlying scientific designs.

• A precise backtesting time span: set from March 10, 1999 to October 2, 2025, a range that deliberately covers the most brutal 2000-to-2002 dot-com bubble in the NASDAQ-100’s history (a drop reaching 83%), the ultimate touchstone for a defensive allocation.
• Precise mapping of Taiwan-listed ETFs: 00662 corresponds to QQQ; 00865B corresponds to BIL; 00670L corresponds to QLD, and for the gap before QLD’s 2006 listing, a daily-compounded back-fill of “NASDAQ-100 daily return × 2, minus an annualized 2.5% financing cost” is used.
• The Moving Block Bootstrap: the simulation draws complete historical blocks of 63 consecutive trading days and stitches them together, which ensures that the simulation paths contain the strangling effect of “a big drop followed by extreme volatility” found in real markets, preserving the “volatility clustering” characteristic of real markets and making the results far closer to reality than random sampling.

The Survival Comparison Under Extreme Scenarios

The simulation data shows that the 433 allocation displayed astonishing survival resilience. Under the harsh conditions of drawing 2% living expenses and requiring a maintenance ratio above 167%, the precise success-rate differences of the three allocations are as follows:

On the surface, the difference in success rate among the three is only about 2%, but the underlying principles are completely different. The 433 allocation, because its 30% leverage component possesses powerful “explosive recovery force,” can rebound at a faster speed when the market recovers, offsetting the fixed annual pledge interest and the 2% drawdown pressure. The 613 allocation, because its leverage is too low, cannot rebound quickly in a long bear market, and its net asset value is continuously eaten away by interest and drawdowns, ultimately falling into the trap of “chronic bleeding.”

This reveals the core of scientific allocation: “recovery capability” matters more than “low leverage.”

Full-Cycle Return Performance

Beyond defensive strength, the 433 allocation also displays a strong growth slope in a normal market. According to the statistics of the simulation report:

• Normal scenario (P50 median): the median assets of the 433 allocation after 26 years are far higher than low-leverage combinations such as 613 or 703. This means that in most historical paths, the coordination of 30% moderate leverage with rebalancing can stably produce an absolute dollar output that beats the broad market.
• Excellent scenario (P95): in bull-market paths where the market rises over the long term, 433’s cumulative total assets display explosive compounding growth, its asset scale even reaching dozens of times the initial investment — this is precisely the practical embodiment of “letting compounding fly.”

If the market drops 85% or 90%, your leveraged fund will never come back in your lifetime. If you do not rebalance, everything you earned in the past is reduced to nothing. (Video 00552)

▲ Figure 11-2 The Monte Carlo P5 extreme stress-test results — under the darkest path, 433 still holds the survival line of defense

A Defense Built on Science

Through the quantitative research provided by Chengfeng Linghang, we can foresee that a correct asset allocation is not about avoiding leverage, but about using a reasonable cash cushion and a precisely calculated Beta to still maintain an absolute survival line of defense above 130% in the darkest moments.

Why is the cash position recommended to be more than 2.2× the pledged borrowing amount? There is a scientific calculation for this: suppose you borrow NT$10,000 at a 5% rate, and after 10 years the principal plus interest is about NT$16,000; if the collateral can be borrowed against at 60%, you back out a need for NT$27,000 in assets; estimating conservatively at a 2% annualized return on the collateral, you back out a present need for NT$22,000 in assets. NT$22,000 divided by the NT$10,000 borrowed is exactly 2.2×. This is the scientific survival talisman for getting through a bear cycle as long as 10 years.

Decoding the Seven Major Scenarios of the 25-Year Historical Backtest

Monte Carlo is a “hypothetical scenario,” but 25 years of real history from 1999 to 2024, covering the dot-com bubble and the financial tsunami, is the most brutal judge. The core dividing line of the scenarios is only one: those “still employed” have a salary to supplement them and do not need to withdraw from the stock market; the “retired” rely entirely on their assets to live, and every decision is a life-or-death choice.

The most shocking comparison comes from the divergent fates of two retirees. Scenario 3.3 (433 allocation + pledged borrowing + smart rebalancing) has a 25-year surplus multiple as high as 5.75×; while Scenario 7 (holding 100% QQQ, living entirely off stock sales) has a 25-year surplus multiple of only 0.92× — a real loss, with an annualized return of -0.35%. Drawing down by selling stock manufactures a permanent capital loss in bear-market years; this is not theory, but the iron verdict handed down by history.

▲ Figure 11-3 A comparison of the surplus multiples of the seven major scenarios in the 25-year historical backtest — allocation structure and rebalancing discipline decide retirement survival

The “comparison of the surplus multiples of the seven major scenarios” gives not an abstract preference, but ordered evidence of retirement survival rates: facing the same drawdown need, whether you can lift the surplus multiple from 0.92× to 5.75× comes down to allocation structure and rebalancing discipline.

The Extreme Backtest Difference Between the Employed and the Retired

To truly understand why the “employed” and the “retired” need different allocation strategies, the most direct way is to compare the asset surplus multiples of the two over the same stretch of extreme history. The following data is based on the actual backtest in Chengfeng Linghang’s EP04 lecture notes, from March 1999 to December 2024 (covering two major bear markets). (Chengfeng Linghang, “EP04”)

Because the employed have a steady salary continuously flowing in and do not need to withdraw from the stock market, they can choose a more aggressive allocation:

• Beta 1.2 (442 allocation): a 25-year asset accumulation multiple of about 13.65× to 14.26×
• Beta 1.0 (433 allocation): a 25-year asset accumulation multiple of about 10.64× to 11.85×

The employed can choose “brainless rebalancing” or even “no rebalancing at all” and still pass safely, with performance maintained above a double-digit multiple. In an environment supplemented by steady cash flow, a high-Beta allocation can indeed bring a markedly higher absolute return.

Once you enter the retirement period and start drawing down, the situation changes fundamentally:

• Beta 1.2 (442) with drawdown: a surplus multiple of about 11.93× to 13.14×, but under the continuous 2000-to-2008 bear market, if smart rebalancing and a sufficient cash buffer are lacking, blow-up risk appears.
• Beta 1.0 (433) with drawdown + smart rebalancing: a surplus multiple of about 9.52× to 10.73×, but with the highest success rate and the lowest psychological pressure. Quantitative data shows that when executing rebalancing in up years, moving “30%” of the profit into the cash position gives a long-term CAGR performance superior to a combination that moves 50% — profit should be left running on the field as much as possible.

This is also why “the retired with no income uniformly adopt the 433 allocation paired with smart rebalancing, with no exceptions allowed.” The 3-to-4× absolute return that is given up buys the certainty of never being forced into liquidation for a lifetime, and this trade far exceeds the numerical gap in value.

For those still employed and with a steady salary, 442 can serve as an advanced choice to accelerate accumulation; but the moment you enter retirement, or need to rely on drawing down assets to live, the allocation should immediately drop back to 433 or a more conservative formation. The core principle is not to “pursue the highest slope,” but that “the capital chain must not break in any year.”

One Hundred Thousand Simulations Upgrade the 4/3/3-to-10/0/0 Stable Zone

Twenty thousand simulations have already given the direction; in 2026 an advanced backtest shared by 00662 community member Pon further pushed the sample to 100,000 runs and expanded the allocation range to 21 variants from 0/5/5 to 10/0/0, comprehensively comparing the long-term performance of “smart rebalancing” versus “brainless rebalancing.” Simulation conditions: the three tickers QQQ / QLD / SGOV, Beta 1.0, a pledged drawdown of 2%, a fixed pledge rate of 7%, annual inflation of 3%, a maintenance-ratio floor of 130%, and 30 years of retirement.

The stable zone of smart rebalancing: the success rates between 4/3/3 and 10/0/0 all land around 93%, with almost no difference in performance. That is to say, as long as leverage is not excessive (not exceeding 30%) and cash and underlying are reasonably distributed, the survival rate of this system is already saturated — further lowering leverage and raising cash yields extremely small marginal benefit.

The cost of brainless rebalancing: although the brainless version’s median annualized return is close to 15% starting from 0/5/5, seemingly stable, its success rate clearly lags the smart version, chasing from 79% at 0/5/5 all the way to 10/0/0 before catching up. In other words: the returns look about the same, but the survival probability is eaten away by 5 to 10 percentage points over the long run — this is the tax that must be paid for “mechanically rebalancing regardless of rises or falls.”

▲ Figure 11-4a The 30-year retirement success rate of the 100,000 simulations (smart vs. brainless rebalancing) — smart rebalancing runs at about 93% between 4/3/3 and 10/0/0, leading the brainless version across the board

Source: 00662 community member Pon, 2026 Monte Carlo 100K simulation

▲ Figure 11-4b The median annualized return of the 100,000 simulations (smart vs. brainless rebalancing) — the two have close median returns; the gap actually hides in the survival rate rather than the return

Source: 00662 community member Pon, 2026 Monte Carlo 100K simulation

Observing the year of failure: beyond the success rate, “in which year the failure occurs” is also a key indicator. The higher the cash position, the later the year of failure appears: 0/5/5 (all-in cash-heavy), if it fails, triggers at about year 15, while 4/3/3 is pushed back to around year 19. Postponing failure equals postponing the day of the blow-up, giving the family 4 to 5 more years of buffer time. From this angle, the stretch between 5/2.5/2.5 and 4/3/3 is the optimal segment that balances both the success rate and “postponing the failure point.”

▲ Figure 11-5 Failure rate vs. year of failure (smart vs. brainless rebalancing) — the higher the cash, the later the failure point; smart rebalancing beats the brainless version on both indicators

Source: 00662 community member Pon, 2026 Monte Carlo 100K simulation

Note: This section is the 2026 upgraded 100,000-simulation version (maintenance ratio 130%), which is a different stress-test baseline from the earlier Figure 11-2 P5 132%, and the data cannot be mixed.