Precise Beta Control

Risk disclosure: The forward-leveraged ETFs discussed in this chapter, such as 00670L, QLD, and TQQQ, use a daily-reset and compounding mechanism, which carries the following characteristics: (1) volatility is amplified (2) long-term holding readily produces decay (3) the contango cost of rolling futures (4) they are unsuitable for anyone who has not fully understood the risk to hold naked or to accumulate through undisciplined regular contributions (if included in an allocation, they must be paired with an underlying ETF, cash / short-term bonds, and rebalancing discipline). The related content in this chapter is all public historical data and an instructional explanation of the leverage mechanism, and is not an investment solicitation or a buy-or-sell recommendation.

In the pursuit of asset growth, many people are readily drawn in by the astonishing performance of leverage tools in a bull market, yet have never delved into the mathematical mechanism hidden behind them. Returning to the most rational logic, we must not only dissect in depth the decay principle and three physical characteristics of leveraged funds, but also break down how, through the precise calculation of the Beta value, to build an investment portfolio that can advance to attack and retreat to defend.

The Essential Difference Between Margin and Leveraged ETFs

When the general public talks about leverage, most people’s intuition is margin, meaning “borrowing money to buy stocks” to magnify profit. But this system does not use margin; it uses leveraged ETFs. Within this CLEC system architecture, one end is the offensive 2× leverage, and the other end is the defensive cash cushion. A 2× leveraged ETF paired with a cash-like short-term bond allocation presents, overall, a configuration that approximates simply putting everything into the underlying (1×) ETF, yet is superior to going all-in on the underlying ETF with no cash, because it holds the confidence and risk resistance that cash provides. Using leverage is therefore not “betting” on higher returns, but rather using the characteristics of leveraged ETFs and the resilience of cash-like assets to build an asset allocation that can advance to attack and retreat to defend, so as to hedge against the unpredictability of the market.

Financial leverage refers to magnifying principal through borrowing, and this can be further divided into three types, each with a different risk profile:

Product leverage refers to using derivatives such as futures to control a large contract with a small amount of margin, such as the 2× leveraged ETFs 00670L and 00675L. The core feature of this kind of instrument is that the investor does not need to borrow from a bank; the funds are handed directly to the fund company to operate in the market. This means that even in an extreme market, the ordinary investor who buys a leveraged ETF is generally limited in their maximum loss to the principal invested, and will not, as with margin, face a margin call or “end up owing money”; but the fund itself may still be liquidated and terminated because of an overly small scale, an extreme market, or regulatory conditions.

Product leverage can indeed deliver rich returns; however, once the market enters a period of choppiness or a prolonged decline, the so-called volatility drag becomes an invisible killer that devours principal.

The underlying mathematics of volatility decay comes from the following formula:

(1-x)(1+x) = 1-x²

Suppose the market falls 5% on the first day and then rises 5% on the second day. The net result is not a return to the starting point, but a drop to 0.9975 (a loss of 0.25%). This x² loss, which inevitably arises each time the market swings back and forth, is volatility decay.

A more intuitive example is this: if the underlying index falls -10% on the first day and rises +11.1% on the second day, the net value can return exactly to the starting point (100 → 90 → 100). But with 2× leverage, after a -20% drop you need a +25% rise to break even; the break-even threshold is raised, and this is the most central source of decay in a choppy market.

The Three Physical Characteristics of Leveraged ETFs

To make the mathematical principle of leverage decay concrete, we must first understand the operating essence of a 2× leveraged ETF (such as Taiwan’s 00670L or the U.S. QLD). This kind of fund seeks a “daily” gain or loss of twice that of the underlying fund, and this mechanism gives rise to the three physical characteristics of leveraged ETFs:

1. Continuous rises produce positive-side compounding: when the market is in a clear one-directional uptrend, the daily-settlement mechanism produces a positive compounding effect, so that the final total gain far exceeds twice that of the underlying index.

Taking two consecutive days of 10% each as an example, the underlying fund’s cumulative gain is 21%.

1.1 × 1.1 = 1.21

The 2× ETF, however, reaches as high as 44%, far exceeding twice the underlying.

1.2 × 1.2 = 1.44

1. Continuous falls produce negative compounding: in a continuously falling market, because the base shrinks each day, the cumulative total decline turns out to be smaller than twice the underlying index’s decline, which has a slight convergence-protection effect in the early stage of a bear market.

Taking two consecutive days of falling 10% each as an example, the underlying fund’s cumulative decline is 19%.

0.9 × 0.9 = 0.81

The 2× ETF’s decline is 36%, which, though large, is slightly less than twice the 19%.

0.8 × 0.8 = 0.64

1. Sideways-choppy volatility decay: this is the drawback of leveraged products. If the index rises 5% and then falls 4.76% back to the starting point, the underlying ETF has lost nothing, but the 2× ETF, due to compounding loss, loses 0.5%. The real-world data from 2024 coldly proves this point: the NASDAQ-100 index (in U.S. dollars) rose about 24.9% for the full year of 2024, but 00670L (2×), a NT-dollar-hedged product tracking the same U.S.-dollar index, gained only 34.93%. The blind spot is this: across a full year of choppy rising, the 2× ETF’s gain did not linearly reach twice the underlying (about 49.8%), and this is the real-world physical manifestation of the volatility decay against which this system repeatedly warns. Strict rebalancing discipline is the only rein for reining it in.

▲ Figure 10-1 Empirical evidence of the 2× ETF's volatility decay (2024), where even in a big bull market the 2× ETF's gain fails to reach a linear 2× of the underlying, the difference being the volatility decay ground away

Taking a rise of 10% followed by a fall of 10% as an example, the underlying fund loses only 1%.

1.1 × 0.9 = 0.99

The 2× ETF, however, loses 4%. This gap will keep accumulating over long-term repeated choppiness, forming a volatility decay that is difficult to make up.

1.2 × 0.8 = 0.96

This phenomenon is not merely a coincidence of a single year in Taiwan’s market; academia established quantitative evidence long ago. Trainor and Baryla, in their 2008 study, used thirty thousand sets of ten-year Monte Carlo simulations to confirm that a typical 2× leveraged ETF, over the long run, magnifies the median annualized return only about 1.4× that of the index (rather than 2×), yet the volatility risk still remains at 2×, and the magnitude of an extreme decline can even be magnified to 4×. What holding a naked 2× ETF buys is the asymmetric disadvantage of “return discounted, risk counted in full” — and this is precisely the mathematical grounding by which the CLEC system strictly prohibits full-position leverage and insists on taming leverage with rebalancing discipline and a cash defensive line.

▲ Figure 10-2 A comparison diagram of the three physical characteristics of leveraged ETFs (positive / negative / sideways decay)

If you want to look solely at “sideways decay,” the most frequently overlooked risk, look directly at the volatility decay diagram: it breaks the difference in net value after an equal rise and fall down into a visualized result.

▲ Figure 10-3 A diagram of the volatility decay of leveraged funds

The meaning of the volatility decay diagram is to break down “seemingly returning to the starting point” into “actually already bleeding.” This set of data reveals the true face of leverage tools under different scenarios. During continuous rises, positive compounding lets QQQ grow +21%, while TQQQ can reach +69%; during continuous falls, even though TQQQ is slightly below three times the decline, the destructive power is still extremely heavy. What must be watched for most is the choppy market: when QQQ suffers only a slight decay, the decay of 3× leverage is markedly magnified, and the threshold for long-term holding is far higher than that of the underlying asset.

However, to state volatility decay directly as “you cannot touch the 2× ETF” is to overcorrect. The key is to see whether the decay or the leverage bonus is larger. The decay does indeed exist, but for a broad market that trends upward over the long run and only swings along the way, the bonus accumulated by 2× leverage during bull segments can, over the long run, usually cover the decay ground away during choppy periods; the 2× ETF as a tool is itself reasonable, and not a reckless bet.

What truly decides life or death has never been “how many times leverage was used,” but “how this 2× is to be allocated.” With the same overall Beta controlled near 1, differing internal makeup of the allocation makes the survival rate worlds apart. This point the author of this book personally verified with a Portfolio Visualizer backtest of 2016 through 2025 with annual rebalancing: the difference is not how many times the leverage is, but “whether there is an underlying as a base.” A configuration like “50% QLD + 50% cash,” which presses half the assets into the 2× ETF while having no underlying at all (that is, the 055 configuration that CLEC explicitly prohibits), has a downside capture rate exceeding 103% and a maximum drawdown as deep as -29.41%, almost as miserable as going all-in on QQQ; and every time a bear market arrives, rebalancing forces you to pour precious cash, one lump after another, into the crashing QLD, making cash depletion very easy. Teacher James therefore condemned it outright as “Russian roulette” and listed it as a prohibited strategy. What can genuinely balance return and survival under a Beta of roughly 1 is the 433, which retains 40% underlying: its downside capture rate is closest to 100%, its maximum drawdown is best controlled, and its cash is sufficient to execute rebalancing at the lows. (Blaise, “Leveraged ETF Allocation: Backtest Data and Psychological Analysis of TQQQ33, 50:50, and 433”)

Behind this lies an iron premise — the asymmetry of gains and losses: the 2× ETF’s advantage of “falling less” can only be obtained when it is paired with cash and adds to positions at the lows through rebalancing; holding a naked 2× ETF into a once-in-a-century crash turns the asymmetry, in an instant, from an advantage into a permanent, drowning catastrophe.

▲ Figure 10-4 A comparison of the compounding and decay of QQQ / QLD / TQQQ under different market scenarios, where in a choppy market the higher the leverage multiple, the more markedly the decay is magnified

The Bankruptcy Backtest Reveals Why Full-Position Leverage Is Strictly Prohibited

We must recognize that although long-term simulation data show that over the 20 years from 2006 to 2026 the U.S. QQQ returned 1,798.57% while the 2× leveraged QLD returned 6,287.87%, this absolutely does not mean you can buy it mindlessly.

Holding a leveraged fund 100% in a single lump position is extremely risky. A 100% leveraged fund still had not broken even after 20 years! Even to this day, 23 years on, it is only one-third of the underlying fund. (Video 00531)

An even more concrete liquidation red line must be flagged: if the market encounters a one-directional extreme crash with a decline reaching 40% or more, a leveraged ETF (especially a 3× leverage such as TQQQ) faces not merely the evaporation of net value, but “direct liquidation and delisting.”

You cannot buy only a leveraged ETF… If the market falls 40% or more, a leveraged ETF will go straight to zero and delist. Buy an ordinary index fund, and no matter how much the market falls, the assets will not go to zero, and they will rise back; but a leveraged ETF will go to zero, and you will go bankrupt. (Video 00164)

What needs to be clarified is that this passage comes from an early stage of CLEC, before the 433 / Beta allocation framework had been developed — at that time it was an “all or nothing” binary judgment. So this absolute warning of “you cannot buy a leveraged ETF” is aimed at “full-position leverage holders with no cash protection” and “3× leverage (TQQQ) holders,” and is not a negation of the 433 / 442 configurations, which took shape only later and are protected by a 30% cash cushion. Under the 433 system, the 2× leverage (QLD / 00670L) accounts for only 30% to 40%, with the remaining 60% to 70% composed of underlying and cash. Even if the 2× leverage position plunges deeply, the 60% to 70% base composed of underlying and cash can still hold the overall portfolio alive; but if you bet 100% of a full position on a leveraged fund, encountering a deep crash will trap you for years in a hard-to-recover position because of volatility decay, and may even force you to sell at a loss at the lows, locking in a permanent loss. An underlying index fund can rise back relatively quickly after a crash, whereas full-position leverage that lacks an underlying base may never make up the difference again. This is the mathematical basis for the CLEC system’s strict prohibition on full-position leverage and its requirement that cash and the underlying serve as an “undying lifeline.”

The Pure-2×-ETF Trap of Holding Cash but 0% Underlying

If the prohibition on “full-position leverage” comes from an early binary judgment, then the next red line is a more refined conclusion converged upon only after the 433 / Beta framework had matured and been verified by backtesting: even if you already know to keep cash and no longer go all-in on leverage, as long as you swap the entire underlying block for the 2× ETF (0% underlying), you will likewise expose a fatal weakness in the backtest. Many people think “I am safe as long as I keep cash,” and so they design a “0% underlying, pure 2× ETF + cash” configuration — for example

• the 055 configuration (50% 2× ETF + 50% cash)
• the 064 configuration (60% 2× ETF + 40% cash)

This is a serious misunderstanding, and it is also the configuration that Teacher James has repeatedly and sternly warned against as absolutely impermissible.

The leveraged-fund position cannot be too large; it cannot exceed the underlying-fund position, nor can it exceed the cash position — this must absolutely be observed. Otherwise, by the time you run into trouble and cry for help, it may already be too late. (Video 00541)

We do not recommend holding a large amount of leveraged funds at all; they are only used paired with cash, and your leveraged fund should not exceed the cash position. This has been said again and again and again. The first thing you must have is stocks, is an index fund. (Video 00531)

Two iron rules must be carved into the mind: “the 2× position cannot exceed the underlying position,” and “the 2× position cannot exceed the cash position.” By contrast, the pure-2×-ETF configurations cross every line:

The first layer of damage is the “recovery trap after a deep fall,” but we must first dispel a misunderstanding that leads people astray. There are two extreme claims about the 2× ETF in circulation: one is alarmist, saying it “will delist and go to zero on a single day’s crash,” and the other, in rebuttal, says “the 2× ETF simply will not delist, so go ahead and hold a large amount of it.” Both are off.

The fact is: the 2× ETF almost never goes to zero and delists — “2× leverage will not go to zero” is correct, but “will not go to zero” is absolutely not the same as “can rise back.” If the market falls 85%, the 2× ETF may fall 99%, with 100 dollars left as only 1 dollar; and to rise from 1 dollar back to 100 dollars requires a full 100-fold rise (+9,900%), a path that is long and hard. So “will not go to zero” means only “will not fall literally to 0 dollars”; it is absolutely not the same as safe, and even less does it mean it can easily climb back after the fall.

Returning to the matter of delisting, the termination threshold for Taiwan’s leveraged ETFs looks at “the fund’s scale being below the statutory floor over the long term,” rather than how much it fell on a particular day. Taking the Fubon NASDAQ-100 2× (00670L) as an example, the termination threshold in its trust contract is a fund net asset below NT$300 million, while its scale has been in the tens of billions for years; even a single-year drop of 62% in 2022 came nowhere close to it (the following year, 2023, it rebounded over 100%; the above is per the prospectus, with the latest official announcement to be taken as authoritative).

But “will not delist” is absolutely not the same as “can go pure 2× ETF.” The true destructive power of the 2× ETF lies not in going to zero, but in the volatility decay and ultra-long trapped period after a deep fall that makes “you doubt your life”: the time it needs to break even is far longer than the underlying’s (in 2008 QLD once hit -83% and took about 5 years to climb back, while the underlying repaired far faster).

The danger hides in this trapped period — should the maintenance ratio go critical during it, or should you need to withdraw living expenses, you will be forced to sell at the lows, cementing paper losses into permanent losses. The 2× ETF will not suffer “sudden death,” but it will slowly drag down a portfolio that lacks an underlying base through prolonged decay. Do not be misled by an equivalence formula such as “50% 2× ETF + 50% cash, with returns roughly equal to 100% underlying” — the returns may be equivalent, but survival absolutely cannot be equivalent.

The second layer of damage hides in rebalancing. What most people default to is “mindless rebalancing” — mechanically resetting the proportions back to fixed values every year. Applied to a pure 2× ETF + cash, this becomes a fatal accelerator in consecutive down years (such as 2000 through 2002, when the NASDAQ-100 fell more than 30% each year for three straight years): the 2× ETF plunges all the way, its weight shrinks, and mindless rebalancing, in order to fill the proportion back up, forces you to pour cash into the still-crashing 2× ETF again and again, which amounts to using your life-saving ammunition to catch a falling knife. When there is no underlying base and the entire stock position is a 2× ETF that decays as it falls, this “buy more the more it falls” lets the net value be rapidly whittled down during consecutive down years, and the cash cushion hits bottom early — often the portfolio is hollowed out before the market can rebound.

In the 433 / 442, this damage is steadily diluted by the 40% underlying: the underlying, after falling deep, can ultimately rise back in full and will not decay as it falls, which amounts to leaving a “recoverable” safe foundation for rebalancing. But in a pure 2× ETF configuration, every dollar that rebalancing adds is thrown into a position that is still decaying — it does not dilute the damage, it directly gnaws at your principal.

The conclusion is one uncompromising discipline: no matter how much cash you keep, the portfolio must have an underlying base “no smaller than the 2× ETF.” This is precisely the fundamental reason why the CLEC standard configuration is always 433 / 442 (retaining 40% underlying), rather than swapping the entire underlying block for the 2× ETF.

The Devastating Drawdown History of the 3× Leveraged TQQQ

▲ Figure 10-5 The devastating drawdown of TQQQ in a back-projected simulation (1999-2009), where 3× leverage was once nearly wiped out and almost lost the ability to recover after the fall

If we push the time back to an extreme bear market, the crash magnitude of a leveraged ETF is extremely staggering. During the dot-com bubble of 2000, the 3× leveraged TQQQ once showed a horrifying decline of as much as 99.96%, crashing all the way from nearly 3,510 dollars to about 1.13 dollars in 2003; while the 2× leveraged QLD’s estimated decline was as deep as -75%, and it took as long as 14 years to return to its prior high.

During the 2008 financial crisis, QLD’s maximum decline was about -83%, and it took 5 years to recover, not rebounding until 2013; TQQQ, meanwhile, once bottomed out at 0.12 U.S. dollars. Even in the recent bear market of 2021 to 2022, when QQQ pulled back about 30% from its highest point, TQQQ’s price also plunged all the way from 88 U.S. dollars to 16 U.S. dollars.

These real-world historical data reveal a cruel reality: after a -99% fall, a 100-fold rise is needed to break even, and this kind of extreme loss triggers a physiological “decision paralysis” in humans, leading to a total loss of the ability to recover. Taking the 2022 crash as an example, some time after the bottoming-out rebound, when the underlying QQQ was down only -6%, QLD was still struggling at -27%, and TQQQ was even still down -44%.

One netizen once shared a real-life tragedy: because he was fully loaded on pledging, when the Trump tariff event hit and 20% of his market value fell away in just 3 days, the maintenance ratio immediately sounded the alarm. This is why full-position leverage is strictly prohibited, and why you must leave a life-saving cash air tank, so that when a black swan descends you can ensure you remain in the game.

Strictly Prohibited Homemade Beta of “TQQQ + a Large Amount of Cash”

We must give a special warning about the “TQQQ + a large amount of cash” homemade combination that beginners most easily misuse. Many beginners think that “1/3 TQQQ + 2/3 cash” can bring the total Beta to roughly 1.0 (1/3 × 3 ≈ 1) while retaining a large amount of cash to add on dips during a decline, seemingly the perfect formation that can “advance to attack and retreat to defend.” Teacher James gives an extremely stern prohibition on this:

Regarding the 1/3 TQQQ + 2/3 cash allocation you are currently adopting, I must say very clearly that this is an allocation method that we repeatedly and sternly prohibit in the CLEC investment philosophy. This is not risk control; it is a mistaken way of using leverage. The premise for rebalancing to be effective is that the asset itself can recover, and after TQQQ falls 90%, it needs to rise 900% to return to the starting point — this is not resilience, it is a hopeless besieged fight, and once you start to average down again and again, you are in fact pouring more and more of your assets into a bottomless pit. (Video 00490A)

The key issue is the “premise for rebalancing to be effective”: adding on dips through rebalancing works on QQQ / QLD because these instruments possess the repair capability of “being able to rise back after falling deep.” But TQQQ’s 3× leverage changes the repair math completely — a 50% fall needs a 100% rise to break even, a 90% fall needs a 900% rise to break even, and a 99% fall needs a 9,900% rise (that is, 100-fold) to break even. When the investor, following rebalancing discipline, adds the 2/3 cash into TQQQ in batches, it amounts to pouring life-saving ammunition, in a steady stream, into a bottomless pit that “mathematically cannot come back.” Each averaging-down makes TQQQ an ever-larger proportion of total assets, until at last all the cash has been converted into the leverage position and the entire account is completely emptied.

The terrifying thing about this trap is that “every step of execution looks like discipline” — the investor is operating according to the “rebalancing bible,” yet does not know that this rebalancing mechanism simply fails on TQQQ. The CLEC system permits QLD (2×) as the mainstay leverage, but toward TQQQ it uniformly and strictly prohibits full positions and strictly prohibits pairing it with a large amount of cash to make “homemade Beta”; this red line cannot be crossed.

The greatest purpose of a leveraged fund is to prepare cash flow and reduce risk, not simply to boost performance! (Video 00496)

Breaking Down the Accumulated Cost of a Lost Decade

The “bankruptcy backtest” above reveals a one-time destruction from an extreme crash, but a more insidious killer is the slow bleeding of a “sideways decade.” When the market enters a long-term range-bound choppiness (such as the NASDAQ from 2000 to 2010), a leveraged ETF’s internal expenses, futures-rolling cost, hedging cost, and volatility decay act like a “reverse compounding,” slicing a piece off the net value every year, and over ten years devour the accumulated gains of the bull period entirely. In a “lost period” where the underlying index has not fallen much, or has even marked time in place, this is precisely the invisible battlefield where leverage holders most easily perish.

To quantify the feel of this “invisible tax,” the annual cost of leveraged ETFs is broken down as follows (using an underlying index annualized volatility of 25% and a sideways decade as the scenario baseline):

The way to read this table is not “you will definitely lose this much every year,” but “in the scenario where the underlying index marks time in place for ten years, how much the leveraged fund will quietly lose.” In other words, in a sideways market, the underlying ETF holder at most breaks even, but the leverage holder will, on the premise that “no crash has occurred at all,” see net value halved or worse.

This also explains why CLEC strictly prohibits “holding a full position of 3× leverage such as TQQQ” — once a lost decade sets in, the invisible tax rate of 3× leverage is as high as 22%, equivalent to being levied about a quarter of your principal by the market every year, a mathematical black hole that any active income struggles to counter. This simultaneously explains why a leveraged fund must be paired with rebalancing and a cash position: rebalancing lets the investor bank leverage gains for safety at the tail end of each bull market, while the cash position continuously hedges the invisible tax during sideways years, keeping the portfolio from falling into the predicament of “the broad market clearly did not fall, yet I keep losing.”

Note: The volatility decay in the table uses the approximation formula “for a 2× ETF in a market with annualized volatility σ, the annual loss ≈ σ²; for a 3× ETF it is 3σ².” Actual decay will vary with the daily volatility distribution, rebalancing frequency, and degree of gapping; this table serves only as an order-of-magnitude conceptual illustration.

The Dynamic Control Formula for Portfolio Beta

Since leveraged funds carry decay risk, we must precisely control volatility through the Beta value. Set the Beta baseline of QQQ (the underlying index fund) at 1.0, and use this as the basis for calculating the volatility of the overall assets:

• Cash / short-term U.S. Treasuries (BOXX/SGOV/00865B) Beta ≈ 0.0
• Underlying fund (QQQ/00662) Beta ≈ 1.0
• 2× leveraged fund (QLD/00670L) Beta ≈ 2.0

Calculate the overall exposure by multiplying “asset proportion” by “asset Beta.” For example, if a portfolio contains 80% underlying fund, 10% 2× leveraged fund, and 10% cash, its total Beta is calculated as follows:

• Total Beta calculation

(0.8 × 1.0) + (0.1 × 2.0) + (0.1 × 0.0) = 0.8 + 0.2 + 0 = 1.0

Through this mathematical logic, the Beta exposure under different configurations can be calculated precisely:

Once these Beta calculation methods are understood, one can grasp how to control the overall volatility of an investment portfolio by adjusting the proportions of different assets — and this is precisely the mathematical basis for translating different investment goals and life stages into concrete configuration combinations.

Given the same Beta of 1.0, why not choose the “50:50 configuration” (50% QLD + 50% cash)? The data show that although the 50:50 configuration has a surface Beta of 1, its downside capture rate exceeds 103%, amplifying the decline effect, with a maximum drawdown reaching -29.41%. If a bear market arrives, each rebalancing requires spending cash to buy the crashing QLD, easily leading to cash depletion and psychological collapse.

The deeper fatal flaw is this: the 50:50 configuration completely abandons the core position in the underlying ETF (00662 / QQQ), which means the investor will in the future be unable to enjoy the CLEC system’s most crucial “securities pledging” advantage. The underlying ETF is relatively stable in volatility, has a low pledging rate (such as Yuanta Securities Finance’s roughly 2.89% for 00662), and is less affected in its maintenance ratio; by contrast, simply holding a 2× leveraged ETF carries a markedly higher pledging rate and violent volatility, and in a big crash very easily triggers a maintenance-ratio margin call. In other words, the 50:50’s surface performance may not be bad, but the cost is giving up “living off borrowed money,” the most precious cash-flow channel in the retirement stage — and this is the true strategic reason the CLEC system insists that the underlying position must be retained.

▲ Figure 10-6 A Monte Carlo simulation comparison of the 433 golden configuration (40/30/30) and the QQQ underlying, where the 433's drawdown is markedly shallower than the underlying's and its recovery burst after a crash is stronger

It is worth noting that the Monte Carlo simulation report indicates that in an extreme-crash stress test paired with pledged withdrawals, the “433 configuration’s” lowest maintenance ratio (132%) is in fact safer than the “613 configuration’s” (119%). From the maximum drawdown (MDD) data on the right side of Figure 10-6, one can see that the “433 configuration,” across a 25-year test, has the bottom of its drawdown box plot markedly higher than the QQQ underlying. This proves that a moderate Beta leverage paired with ample cash lets assets possess a stronger “recovery burst” after a crash.

Teacher James’s core system emphasizes controlling overall Beta risk through the cash proportion (keeping Beta around 0.7 to 1.0) rather than over-leveraging, so as to ensure the long-term survival of the system.

The Decline Tables of QQQ and QLD Across Various Crash Tiers

The most direct way to understand the risk of a leveraged fund is to quantify its true decay across crashes of different magnitudes. The following data reveal the reality of the declines one must be psychologically prepared to bear when holding the 2× leveraged QLD.

Facing this kind of extreme scenario, aside from allocating cash, “salary as living water” is the last line of defense. A calculation shows that in a -80% devastating crash, if no funds are added at all, the pledged portfolio’s maintenance ratio would fall below the 167% safety line; but if NT$10,000 of salary is continuously injected each month, the maintenance ratio can hold steady at 312%. This proves that the deeper the crash, the thicker the cash position and salary living water must be.

Chengfeng Linghang also points out the “exchange-rate buffer” effect that NT-dollar investors may enjoy: in certain risk events, a sharp fall in U.S. stocks is often accompanied by NT-dollar depreciation against the U.S. dollar, partly buffering the decline of unhedged NT-dollar products (a scenario illustration, not a fixed rule: for example, when U.S. stocks fall 10% and the NT dollar depreciates about 2% at the same time, 00662’s NT-dollar decline is about 8.2%; the actual magnitude depends on the exchange-rate movement at the time). But 00670L, because it bears the leveraged decline, falls about 21%, equivalent to 2.5× that of 00662. This exchange-rate protection mechanism has a marked effect before a “40% decline,” but once QQQ’s decline exceeds 40%, 00670L’s crash speed will far exceed twofold. For example, when QQQ falls 50% and the NT dollar depreciates 12%, 00662 falls 44%, while 00670L may plunge 88% (this is an extreme scenario overlaid with NT-dollar depreciation, hence deeper than the pure-leverage decline of -78% to -85% in this chapter’s own table). (Chengfeng Linghang, “EP06 Ultimate Simplicity: Mindless Adding-On, Making Something from Nothing”)

This reveals the absolute necessity of retaining a “cash cushion” in the portfolio. In addition, when investing in a Taiwan-version 2× ETF linked to U.S. stocks (such as 00670L), one should not demand an exact “precise 2× every day” in the short term, because the NT dollar’s exchange-rate fluctuation in the 28 to 33 range produces short-term interference; but historical backtests indicate that when the time is stretched to 20 years, the exchange-rate impact will ultimately tend toward neutrality.

Futures Spread Erodes the Net Value of Leveraged ETFs

The fourth physical characteristic of a leveraged ETF is the “futures-spread convergence effect,” which is rarely given weight by ordinary investors. Understanding this mechanism will help in making a shrewder choice of instrument.

Contango is the hidden cost of leveraged ETFs, such as 00670L, that mainly replicate through futures. The U.S. futures market is normally in a state where the futures price is higher than the spot price, and at each “roll over,” the ETF must sell the cheap contract about to expire and buy the more expensive next-period contract, which amounts to “selling low and buying high.” On top of this, because 00670L maintains an NT-dollar position, it loses the high-interest spread of the U.S. dollar and must bear a foreign-currency hedging cost (about 0.6%) and internal expenses, so its real annualized holding cost may be as high as 3.73%, and long-term accumulation will seriously erode the net value.

Laying out 00670L’s decay structure completely will be even more startling. 00670L is not a “contract for difference (CFD)” of the foreign-exchange market, but replicates twice the broad market’s daily gain or loss by buying “NASDAQ-100 index futures contracts” with 200% exposure. Three layers of cost erode the net value simultaneously:

• Futures contango: the hidden bleed of the aforementioned monthly “selling low and buying high” roll-over.
• Borrowing cost: opening 2× leverage essentially requires borrowing money, and in the high-interest environment of 2025, the base rate of U.S. stock futures plus the swap-contract cost lands at roughly 4% to 4.5%; because the own funds can simultaneously be parked in short-term bonds earning interest, only the additional 100% leverage portion is truly financed, and after the offset the annualized net financing drag is about 4.5%.
• FX hedging cost: to keep the net value from being affected by the U.S. dollar’s rise and fall, 00670L must continuously execute foreign-currency hedging, but in a global high-interest environment, the hedging cost itself is very expensive, which further widens the tracking error.

After the three layers of cost stack up, 00670L’s long-term return will produce a significant gap from the U.S.-native QLD.

Another frequently overlooked blind spot is the “currency attribute of the asset.” 00670L uses an exchange-rate hedging mechanism, which substantially reduces the USD/NT-dollar exchange-rate exposure compared with an unhedged product, and is in nature close to an “NT-dollar asset”; but hedging does not mean completely eliminating exchange-rate impact, and there may still be a hedging cost, tracking error, and residual exchange-rate risk. For an investor based in Taiwan who earns an NT-dollar salary to begin with and holds owner-occupied property and a Taiwan-stock position, adding 00670L amounts to pressing even more chips onto the single fiat currency of the NT dollar, so the effect of diversifying geographic risk is instead weakened.

Based on the above three-layer cost and currency-concentration issues, CLEC does not regard 00670L as the “first-choice leverage instrument for long-term holding”; only when an investor has a strong “Taiwan-domestic pledging need” (because sub-brokerage U.S. stocks and overseas brokerages generally cannot serve as pledged collateral within Taiwan) do they need to consider allocating some 00670L as a supporting tool for the cash-flow defensive line. If there is no pledging need, a more direct choice is QLD (U.S.-native 2× leverage) — sparing the three layers of decay and simultaneously obtaining genuine U.S.-dollar asset exposure.

As for the selection logic of Taiwan-domestic 2× ETFs (such as 00631L / 00675L / 00685L), the market-structure shift of Taiwan index futures from backwardation to contango, and practical details such as 00631L beginning to buy the spot to counter the bleed — because these belong to non-CLEC-mainstay content, they are compiled and included together in the appendix section “A Complete Guide to Taiwan-Stock 2× ETFs (Extended Reference).”

Using the System to Withstand Choppiness Decay

Delving deeply into the decay math and three physical characteristics of leveraged funds is meant to keep the investor clear-headed about market risk even in the face of astonishing bull-market returns. Learning to control the Beta value of an investment portfolio is precisely what gives the investor a weapon with which to respond calmly amid the clash of bull and bear.

Many investors who have encountered technical analysis often subscribe to a dogma: “leveraged ETFs are violently volatile, so you can only do right-side trend-following trades; adding on the left side is catching a falling knife.” This statement has its reason when applied to an individual stock — because an individual stock can indeed go bankrupt and to zero. But forcing this logic onto a broad-market leverage tool (such as 00670L, QLD) is imposing the throttle logic of a taxi onto the engine of an airplane.

Behind a broad-market index stands the entire economic output of a nation, carrying its own market-cap-weighted automatic mechanism of culling the weak and keeping the strong; when any single member declines, a stronger enterprise instantly fills the gap. In a long-bull pattern, when the market wrongly kills prices out of panic, this is precisely the moment when quality assets go on a great discount for the capitalist. Holding to left-side discipline and adding in batches to broad-market leverage, what you buy is not an uncatchable knife, but discounted gold.

Cast off the fantasy of full-position leverage; aside from precisely allocating asset proportions, you must build a “decoupling mechanism”: retain at least one year’s worth of personal-loan principal and interest in cash, keeping it completely independent of stock-market volatility. As long as the survival cash flow is not broken, you can sleep soundly when the market collapses, letting time and compounding become the strongest aid in accumulating wealth.