LETF (leveraged ETF)

[1nvest]

Not a fan of HYP myself, prefer the single click type option of buying/selling FTSE250 index exposure instead. Historically the FTSE250 has been a good choice of general/broad index, earns around half of earnings from foreign, is less inclined to see single stocks rise to being 10% of the index, even includes a broad range of Investment Trusts, and could even be considered as including a element of HYP type holdings. Historically the FTSE250 has somewhat aligned with the likes of HYP in terms of total (accumulation) returns.


The charts show yearly total returns (left hand scale), and accumulation total returns (log scaled right hand scale), compared to Terry's (TJH) Accumulation HYP as a benchmark/reference.

Leveraged ETF's typically scale up exposure to a index, the LETF fund managers tend to rebalance daily, and in effect borrow to fund increasing stock exposure. Often snubbed as the tendency is to look at how the LETF worked alone. When however looked at as a means to scale down exposure, such as half as much in a 2x FTSE250 LETF (such as 2MCL) as would have been invested in a standard FTSE250 tracker, then the rewards tend to broadly align/compare. For the other half my personal preference is to hold gold, as I'm not a fan of Gilts/bonds. The last chart in the above image shows the result of 50/50 yearly rebalanced 2MCL/Gold since April 2014 (2MCL inception date was July 2013)

With half as much in a 2x LETF you can generally get away with rebalancing just once yearly. For higher levels of leverage such a as a third in 3x you probably need to rebalance every six months. At more extremes, such how Zvi Bodie holds 10% in 10x (using Traded Options), 90% in 'safe' assets, you'll probably need to rebalance at least monthly. So with a 2x, once yearly rebalanced, then that fits in well with being reviewed/rebalanced at around the end/start of each fiscal year (where that choice of review date enables you to opt to either trade in the old or new financial year (or a combination of both) according to whatever might be the more appropriate at the time - such as for tax efficiencies). With fiscal year end April 2023 rapidly approaching, its looking like 50/50 2MCL/gold is going to end the year around -8% down if things (prices) stay as they recently are.

Just thought I'd post this as there's little in the way of LETF discussions around TLF sections, and as I'm in the process of year end rebalancing/review ahead of the end of 2022/23 fiscal year I had the data/charts to hand.

[1nvest]

3LUS is a 3x US S&P500 leveraged ETF, that trades in London £

Backtesting a yearly rebalanced third weighting to that, along with a third in gold, and a third in a three-year Gilt ladder (fully protected, no matter how much you might 'deposit'), applying a 30 year 4% SWR, and the historic final portfolio value as a multiple of the inflation adjusted start date portfolio looks like ...



In the worst of cases you were still left with around two thirds of the inflation adjusted start date portfolio value after 30 years of 4% SWR (4% of the initial portfolio value drawn in the first year, with that amount uplifted by inflation as the amount drawn in subsequent years, so a nice regular inflation adjusted income stream).

Counter-party risk is considerably reduced, a third in safe Gilts, a third in physical in-hand gold, whilst also being diversified across £/Gilts, US$/stocks (3LUS base currency is US$), and global currency/commodity (gold).

The US version of that PV reasonably closely tracked the S&P500 (total returns).

The only reason I went with a 3 year Gilt ladder rather than T-Bills is that is easier to actually hold. The UK doesn't make T-Bill purchases easy for retail investors to buy, and fully rolling a 1 year Gilt each year entails more costs than if you maintain a 3 year ladder. As another alternative IGLS (a Gilt ETF fund) would likely serve as equally as well

[floyd3592]

Excellent 'food for thought' posts as per usual...
Thanks 1nvest.

[Bubblesofearth]

A word of caution on leveraged ETF's. These products are only really useful for short-term trading opportunities. Longer-term their value will decay with market volatility. This is easily demonstrated with simple maths.

Consider £1000 invested in a 2X leveraged product. Underlaying market falls 10%, product down 20% so values are now;

Underlying £900
2X Leveraged product £800

Market then rises 11%;

Underlying £999
2X Leveraged product £976

Market then falls 10%;

Underlying £899
2X Leveraged product £781

Market then rises 11%;

Underlying £998
2X Leveraged product £953

I've used high percentages to show oscillations in a roughly static market that highlight the decay mentioned. You can play with the numbers but will find that the longer you are exposed to a leveraged product, and the more leveraged it is, the worse will be the decay.

BoE

[minnow]

These products are only really useful for short-term trading opportunities.

Not sure I agree. As long as you're not using silly amounts of leverage, a 2x or 3x levered ETF can be a very sensible long-term hold. This page has a good analysis : http://www.ddnum.com/articles/leveragedETFs.php.

And, of course, you can then use the left-over cash (that you would otherwise have had to lock up in a 1x fund) to purchase a diversifying asset. @1invest likes gold, but I chose 3TYL (3x 10Y US treasuries). The end result is that you can (theoretically) achieve a more efficient portfolio, without having to sacrifice overall returns.

[Bubblesofearth]

Not sure I agree. As long as you're not using silly amounts of leverage, a 2x or 3x levered ETF can be a very sensible long-term hold. This page has a good analysis : http://www.ddnum.com/articles/leveragedETFs.php.

The basic maths stated early in the paper are correct but this statement is not;

The myth has resulted from the belief that volatility drag will drag any leveraged ETF down to zero given enough time.

I did not claim this and the only time I've seen it (correctly) claimed is if the ETF magnifies the market fall on any given day to more than 100% (e.g. a 34% fall of a 3X leveraged product).

Also this statement is misleading;

That’s not fair! The market has gone down by 5% then up by 5% but our ETF that has a leverage of 1 has gone down by 0.25%. Doggone it!


In my example I was careful to compare the effect of oscillations in a flat market on leveraged returns. Leveraged products show decay in the situation where what the paper calls leverage of 1 does not. To give a real and relevant (to many on these boards) example, the capital value of the FTSE 100 has gone virtually nowhere for some 23 years. Direct investment in, for example, a FTSE100 tracker, would have essentially trod water. A leveraged FTSE100 product would not have been so lucky.

Yes, I would agree that the volatility drag associated with a leveraged product would be compensated for given a sufficient rise but I'm not convinced there hasn't been some hindsight bias in the time-frame of the examples given.

Then there is the issue of costs which the paper also acknowledges.

BoE

[minnow]

A leveraged FTSE100 product would not have been so lucky.

Perhaps worth mentioning that that products such as 3UKL don't just track the capital value of the index. They track its total return (from the KID : "The ETP tracks the FTSE 100 Daily Super Leveraged RT TR Index (UKXL3X), providing a total return comprised of 3 times the daily performance of the FTSE 100 Net Dividend TR Index").

Over the past 3 years, the FTSE is up about flat in terms of its capital value, but the total return is somewhere close to 40%. Meanwhile, 3UKL is up 102%. Now of course, I've cherry-picked the starting point to bolster my argument -- but I still reckon there's something to be said for a bit of leverage, if you've got the stomach for it.

[mc2fool]

A leveraged FTSE100 product would not have been so lucky.

Perhaps worth mentioning that that products such as 3UKL don't just track the capital value of the index. They track its total return (from the KID : "The ETP tracks the FTSE 100 Daily Super Leveraged RT TR Index (UKXL3X), providing a total return comprised of 3 times the daily performance of the FTSE 100 Net Dividend TR Index").

Over the past 3 years, the FTSE is up about flat in terms of its capital value, but the total return is somewhere close to 40%. Meanwhile, 3UKL is up 102%. Now of course, I've cherry-picked the starting point to bolster my argument -- but I still reckon there's something to be said for a bit of leverage, if you've got the stomach for it.

3UKL since launch vs CUKX, iShares Core FTSE 100 ETF (Acc)


https://uk.advfn.com/stock-market/london/wt-ftse-100-3x-3UKL/share-chart

[1nvest]

This is yearly rebalanced half in 2x (ULPIX), half in bonds, since 1998, US data (click image to link to the PV data)



You can hold LETF's mid/longer term and see reasonable tracking when utilised to target the same/similar exposure as 1x. Inadvisable however to use them mid to longer term in isolation (leveraging up exposure), as that just tends to scale the volatility whilst broadly yielding similar reward to the non-leveraged.

For a third in 3x, then if the market crashed any more than 33% without you having rebalanced, then you still have the other two-thirds available. Ditto if there were problems with the broker/custodian. With 100% 1x you're exposed to the full downside. In that 3x example PV link I posted earlier, imagine if the stock markets closed (using a extreme/silly example) for a number of months, whilst you held the cash in high street bank deposit accounts and physical in-hand gold that were still both accessible. Counter-party risk reduction via diversification so-to-speak. Same/similar rewards, but more (counter-party) risk diversified, as well as capping the downside between rebalancing (assuming cash and gold to be safe).

But as ever, risk reduction via diversification tends to yield other different risks. Holding physical gold in-hand for instance has no counter-party risk, but has security risk (risk of theft). However concentration risk is a major risk factor, 100% concentration into a single currency, geopolitical risk ...etc. and if that does turn bad then you might lose-all. If you restrict single concentration risk to no more than say 33% in a single currency/broker/country/asset ...etc. then as uncomfortable as a 33% might be, still having 66% and having avoided a total loss has you still at the table.

[1nvest]

Another factor to consider is that indeed held alone and sooner or later a large down in a LETF value will occur. Someone with both SIPP and ISA might be best served by holding the LETF in their SIPP, as when those big-down(s) occur so that will drain the SIPP, whilst in combination the overall portfolio still tends to track the 1x. SIPP's are tax free on the way in, taxable on the way out, ISA's are out of taxed income on the way in, tax free on the way out. If over time you add to SIPP, see migration of funds over to ISA, then you have more that was both tax free on the way in and out. Migration from SIPP to ISA also means that capital that was otherwise inaccessible until a certain age can become accessible.

[1nvest]

Since 1930's ending of money=gold, prior to which generally bonds alone were good enough, yielded similar total returns to stocks (inflation broadly averaged 0% so money (gold) deposited (into bonds) for interest, was like a real rate of return), inflation has in effect just become another form of taxation. Print/spend money, which devalues all other notes in circulation (induces inflation), where combined inflation + taxation negates any real returns that bonds previously paid when on the gold standard.

50/50 US stock/gold for British investors since the 1930's supported a 25 year 4.4% SWR. Enough to see a 65 year old retiree through to age 90, beyond which perhaps their home value covers all-inclusive late life care home costs. For 100% stock, that dropped to a 3.4% SWR. All stock had the higher CAGR total return, but its higher volatility was a risk.

For someone content with 50/50 stock/gold then that can instead be held as 17/83 3x stock/gold, or 25/75 2x stock/gold.

But why would anyone be mad enough to become such a gold-bug? Well maybe ... viewtopic.php?p=578235#p578235 Which might perhaps be considered as a form of Black Swan type event, i.e. punitive domestic taxation/punitive-policies. But as with most hedges, is something that has to be in place beforehand, as more likely at the time (or later) its too late, doors will have be slammed shut and locked.

Some say gold is for the Mad Max scenario, and even then is of little use after total collapse. But all fiat currencies sooner or later collapse towards zero, for instance a 1930's US dollar has lost over 99% of its gold purchase power, in effect the fiat US dollar has lost near all of its value. Gold in contrast as a commodity based currency, has maintained purchase power across millennia, albeit in a broad only average manner, at times with considerable volatility around that. But periodically realigns again. 1946 and US stocks paid 4% dividends, again in 1986 they paid 4% dividends. The price of a S&P500 share was around 11.5 times higher, as was the price of gold (11.3 times higher). Holding gold instead of stocks fundamentally forwent the dividend benefits from holding stocks, similar to how also does holding bonds do similar. Forgoing some of the upside, but reducing volatility such that the downside risk is lower.

[1nvest]

With less than a couple of weeks remaining of the current fiscal year, looking like 10/10/20/30/30 will end the year down around -1.5% if recent prices stayed the same

10% 3LUS = 3x S&P500
10% 3UKL = 3x FT100
20% 2MCL = 2x FT250
... with the remainder 60% split equally between cash and gold (and in recent years with near zero cash interest you might as well stuffed physical gold and hard cash under the mattress)



Might broadly be expected to reflect 100% stock exposure.

Data values shown are from WisdomTree (the ETF provider) for the fiscal NAV values.
Gold sourced from LBMA (3pm fix).
Pound/Dollar rates sourced from FRED

3LUS is the years biggest loser, down around -45%, despite the benefit of US$ having gained around +7%. But at least in other years it pops to the upside (such as +233% in 2020/21). 10% weighted = -4.5%, which was in part negated by gold gaining +10.5% (around +3% when 30% weight). Other than that 'action' everything else was relatively quiet/flat.

[vand]

I don't see the point at all myself, especially in this market where the heightened volatility will blunt the edge of daily rebalanced LETFs and you may only bag a fraction of the upside - leveraged ETFs are reset on a DAILY basis and this makes then incredibly poorly suited to long term long strategies.

The supposedly x3 leveraged TQQQ is down by about 20% since Jan 2020 despite the QQQ being up 30% since then precisely because of this daily resetting phenomenon.

A better option if you want leveraged exposure is just put down a spreadbet and put down the appropriate amount of margin. Again, not something I would advise because using margin tends to make even the otherwise most rational investors do the stupidest and most emotional things.

[1nvest]

I don't see the point at all myself, especially in this market where the heightened volatility will blunt the edge of daily rebalanced LETFs and you may only bag a fraction of the upside - leveraged ETFs are reset on a DAILY basis and this makes then incredibly poorly suited to long term long strategies.

The supposedly x3 leveraged TQQQ is down by about 20% since Jan 2020 despite the QQQ being up 30% since then precisely because of this daily resetting phenomenon.

A better option if you want leveraged exposure is just put down a spreadbet and put down the appropriate amount of margin. Again, not something I would advise because using margin tends to make even the otherwise most rational investors do the stupidest and most emotional things.

Jan 2000 to end of Feb 2023 33/67 TQQQ/CASHX (T-Bills) and $10,000 grew to $14,375, whereas $10,000 in QQQ grew to $14,055

PV

Going back further and from 2011 to recent, $10K in 33/67 TQQQ/CASHX $61,069 vs $60,377 for QQQ


PV

Another benefit is that in holding cash as part of the portfolio you might borrow from that cash. £2000 expense, not paid for another 3 weeks, borrow the £2000 from the cash and return it after pay-day. With 100% stock you'd have to sell shares.

A risk with spreadbets is that their systems might foul up and leave you out of a position at a bad price. Computer glitch, prices spike down very briefly, trigger account positions being closed at those lows, to a few seconds later prices having recovered back to 'normal', but where your position was closed out at a large loss.

[vand]

I don't see the point at all myself, especially in this market where the heightened volatility will blunt the edge of daily rebalanced LETFs and you may only bag a fraction of the upside - leveraged ETFs are reset on a DAILY basis and this makes then incredibly poorly suited to long term long strategies.

The supposedly x3 leveraged TQQQ is down by about 20% since Jan 2020 despite the QQQ being up 30% since then precisely because of this daily resetting phenomenon.

A better option if you want leveraged exposure is just put down a spreadbet and put down the appropriate amount of margin. Again, not something I would advise because using margin tends to make even the otherwise most rational investors do the stupidest and most emotional things.

Jan 2000 to end of Feb 2023 33/67 TQQQ/CASHX (T-Bills) and $10,000 grew to $14,375, whereas $10,000 in QQQ grew to $14,055

PV

Going back further and from 2011 to recent, $10K in 33/67 TQQQ/CASHX $61,069 vs $60,377 for QQQ


PV

Another benefit is that in holding cash as part of the portfolio you might borrow from that cash. £2000 expense, not paid for another 3 weeks, borrow the £2000 from the cash and return it after pay-day. With 100% stock you'd have to sell shares.

A risk with spreadbets is that their systems might foul up and leave you out of a position at a bad price. Computer glitch, prices spike down very briefly, trigger account positions being closed at those lows, to a few seconds later prices having recovered back to 'normal', but where your position was closed out at a large loss.

Er, you're wrong... TQQQ doesn't goes back as far as 2000. If it did you would be down by more than 99% (likely more than 99.9% probably) on the TQQQ position from peak to trough and still be underwater on it today.

Just look at how it translated a 35% loss to an 85% drawdown in TQQQ in 2022 and has only rallied back about 60% so still sits much closer to last years' lows. A 80%+ fall in the underlier similar to dotcom be exponentially worse in TQQQ. It's likely they would probably just wrap up the fund like they have done with other leveraged ETFs (XIV) that suffer huge falls, so in practice you would suffer all of the downside but have no vehicle to ride the upside back even if you could stomach it.

[vand]

The problem with what you're testing is that you have the benefit of cherry picking something that we already know has gone up over time and then backtested it. We know that generally speaking if you can leverage into something that has gone up over the long term it generally works out pretty well. But you don't know what asset classes are going to do what over the next cycle. 33/67 in TQQQ/CASHX and TQQQ may go nowhere for the next decade and just destroy your capital with its continual volatility. Oil is averaging the same price it averaged 10 years ago and yet the USO fund has seen a ~90% loss of capital - would be hard pressed to find a strategy that could fit that in...