Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to johnstevens77,Bhoddhisatva,scotia,Anonymous,Cornytiv34, for Donating to support the site

Lyxor

Index tracking funds and ETFs
tjh290633
Lemon Half
Posts: 8209
Joined: November 4th, 2016, 11:20 am
Has thanked: 913 times
Been thanked: 4097 times

Re: Lyxor

#137389

Postby tjh290633 » May 7th, 2018, 5:33 pm

The point is that all other factors are seldom the same.

Theory and practice are very different.

TJH

OhNoNotimAgain
Lemon Slice
Posts: 767
Joined: November 4th, 2016, 11:51 am
Has thanked: 71 times
Been thanked: 147 times

Re: Lyxor

#137472

Postby OhNoNotimAgain » May 8th, 2018, 10:01 am

tjh290633 wrote:The point is that all other factors are seldom the same.

Theory and practice are very different.

TJH


Indeed, which is what Long Term Capital Management discovered.

GeoffF100
Lemon Quarter
Posts: 4724
Joined: November 14th, 2016, 7:33 pm
Has thanked: 178 times
Been thanked: 1363 times

Re: Lyxor

#137621

Postby GeoffF100 » May 8th, 2018, 7:33 pm

Volatility and liquidity are the main determinants of spread in normal conditions. We do not need complicated mathematics to see this.

Suppose that the daily volatility of a share is 1%. That means that we expect a movement of about 1% in the share price over the next day. Suppose that there are 100 transactions per day. The average time between transactions is 0.01 days. The volatility over the average time between transactions is 1% times the square root of 0.01, i.e. 0.1%. According to the paper referenced earlier, the spread, in normal conditions, will be about 3.5 times this volatility, i.e. 0.35%.

This simple calculation illustrates how the volatility and liquidity combine to give the spread. There are a host of other factors that can affect the spread in practice, but that is the exception rather than the rule.

Spet0789
Lemon Quarter
Posts: 1913
Joined: June 21st, 2017, 12:02 am
Has thanked: 242 times
Been thanked: 952 times

Re: Lyxor

#137652

Postby Spet0789 » May 8th, 2018, 10:47 pm

colin wrote:
Lyxor's more volatile long gilt fund GILS appears to have consistently wider spreads than the short dates GIL5


Perhaps the longer dated bonds are less frequently traded.
Rapid falls in the market are accompanied by increases in volatility and increases in both share and ETF spreads

Yes this is true, but that does not imply a causality between low bid offer spreads and low volatility. And that's where the root of your misunderstanding lies.


No, the reason that the ETF on longer-dated bonds has higher spread is that its price volatility is greater.

Bonds trade as a function of yield. From the yield we calculate the price. For a 0.01% change in yield, the price of a 30yr bond changes by about 0.23%. The price of a 2yr bond changes by 0.02%. This relationship is the (modified) duration of the bond.

And yes, spreads certainly are wider (all else equal) for a more volatile underlying. From a market-maker’s perspective, their fair spread is a function of the time they expect to hold the position (liquidity) x the amount they expect the price may change (volatility).

colin
Lemon Slice
Posts: 663
Joined: December 10th, 2016, 7:16 pm
Has thanked: 24 times
Been thanked: 114 times

Re: Lyxor

#137696

Postby colin » May 9th, 2018, 10:04 am

No, the reason that the ETF on longer-dated bonds has higher spread is that its price volatility is greater.



These are the realities of what has been proved or assumed to have been proved so far
non of the following statements are in any way theoretical

1) That on the most volatile day of trading in recent years according to Geoff100 bid offer spreads fell.
2) Three ETFs that track the S&P 500 which by definition must all share the same volatility all have different bid offer spreads
3) The volatility of a preference share (which is about as long dated as one can get) is lower than the volatility of the same company's ordinary share yet the bid offer spread is higher.

The real world , because it is complicated and subject to feed back loops of changing human behavior cannot be predicted by deterministic equations.

hiriskpaul
Lemon Quarter
Posts: 3852
Joined: November 4th, 2016, 1:04 pm
Has thanked: 682 times
Been thanked: 1489 times

Re: Lyxor

#137723

Postby hiriskpaul » May 9th, 2018, 11:33 am

colin wrote:
No, the reason that the ETF on longer-dated bonds has higher spread is that its price volatility is greater.



These are the realities of what has been proved or assumed to have been proved so far
non of the following statements are in any way theoretical

1) That on the most volatile day of trading in recent years according to Geoff100 bid offer spreads fell.
2) Three ETFs that track the S&P 500 which by definition must all share the same volatility all have different bid offer spreads
3) The volatility of a preference share (which is about as long dated as one can get) is lower than the volatility of the same company's ordinary share yet the bid offer spread is higher.

The real world , because it is complicated and subject to feed back loops of changing human behavior cannot be predicted by deterministic equations.

Deterministic no, probabilistic yes. In some areas though, such as investment grade bond and financial derivative pricing, deterministic equations work well enough, for the simple reason that market participants all use the same equations!

I can confirm that the risk monitoring of MM activity does consider volatility and liquidity in precisely the way that Spet0789 and GeoffF100 have described. In some banks, if a MM's spread differs widely from what is expected (produced by the equations) they need to have a damned good reason for it if they want to keep their jobs.

hiriskpaul
Lemon Quarter
Posts: 3852
Joined: November 4th, 2016, 1:04 pm
Has thanked: 682 times
Been thanked: 1489 times

Re: Lyxor

#137732

Postby hiriskpaul » May 9th, 2018, 11:48 am

By the way, it is not at all unreasonable for ETFs tracking the same index to have different spreads as they will have different liquidity. The largest ETF on the planet is SPY and it trades on a miniscule spread. In terms of management charges it is far from being the cheapest ETF. Despite the higher charges, institutional investors continue to trade SPY partly because it is expected to trade on low spreads (and partly because of out of hours trading, ease of shorting and the active options market).

Edit - preference shares are typically much less liquid than ordinary shares, which is the reason for the higher spreads on prefs.

Spet0789
Lemon Quarter
Posts: 1913
Joined: June 21st, 2017, 12:02 am
Has thanked: 242 times
Been thanked: 952 times

Re: Lyxor

#137942

Postby Spet0789 » May 9th, 2018, 10:37 pm

colin wrote:
No, the reason that the ETF on longer-dated bonds has higher spread is that its price volatility is greater.



These are the realities of what has been proved or assumed to have been proved so far
non of the following statements are in any way theoretical

1) That on the most volatile day of trading in recent years according to Geoff100 bid offer spreads fell.
2) Three ETFs that track the S&P 500 which by definition must all share the same volatility all have different bid offer spreads
3) The volatility of a preference share (which is about as long dated as one can get) is lower than the volatility of the same company's ordinary share yet the bid offer spread is higher.

The real world , because it is complicated and subject to feed back loops of changing human behavior cannot be predicted by deterministic equations.


You seem to have either not read or not understood my point.

Firstly, I said, all else equal. There do exist other factors beyond volatility and volume although they are most important.

Secondly you have ignored the second factor which is volume (itself a linear proxy for expected holding time).

In the example you give of different SP 500 ETFs, yes the volatility will be the same but if the volumes are different then the spreads will likely be different. All else equal, spread will be a function of sqrt(1/volume). I suspect volume also explains 1 and 3.

The other points to bear in mind with ETFs are the precise mechanics of the create/redeem process and whether a sponsor bank is showing an articifically tight bid offer to grow the product.

If you still disagree please feel free to set yourself up in the market-making business. You’ll do well as the others are clearly all doing it wrong.

GeoffF100
Lemon Quarter
Posts: 4724
Joined: November 14th, 2016, 7:33 pm
Has thanked: 178 times
Been thanked: 1363 times

Re: Lyxor

#137980

Postby GeoffF100 » May 10th, 2018, 7:44 am

colin wrote:
No, the reason that the ETF on longer-dated bonds has higher spread is that its price volatility is greater.



These are the realities of what has been proved or assumed to have been proved so far
non of the following statements are in any way theoretical

1) That on the most volatile day of trading in recent years according to Geoff100 bid offer spreads fell.
2) Three ETFs that track the S&P 500 which by definition must all share the same volatility all have different bid offer spreads
3) The volatility of a preference share (which is about as long dated as one can get) is lower than the volatility of the same company's ordinary share yet the bid offer spread is higher.

The real world , because it is complicated and subject to feed back loops of changing human behavior cannot be predicted by deterministic equations.

(1). Not true. You clearly have not read or have not understood the article. It states:

"The day after the referendum, the markets exhibited substantial turmoil, large price jumps, increased volatility and increased spreads. Still, we find that the spread did not increase in line with the increase in volatility, making pure market making a losing strategy. However, the nature of the flow on that day might have compensated for it, which explains why the spread was kept relatively low."

Figure 1 clearly shows the spread on the day increased linearly with the volatility per trade (i.e. volatility times the square root of the average time between trades), but not as steeply as usual. This departure from the usual behaviour that was unexpected enough to prompt the article. If the author is right, one (or more) of the market makers was indulging in risky momentum trading. It would be ironic if the trader concerned made his employer money and lost his job.

(2). As others have said, differences in liquidity are the main cause.

(3). Preference shares have much lower liquidity than ordinary shares in the same market.

The equations are not just theory. They are based on sound statistical principles. Market makers who use them make steady profits with little risk. In most market conditions, the spreads in the market closely match the spreads predicted by the equations.

Cookie
2 Lemon pips
Posts: 120
Joined: January 27th, 2018, 10:19 pm
Has thanked: 63 times
Been thanked: 6 times

Re: Lyxor

#139714

Postby Cookie » May 17th, 2018, 9:54 pm

Well this thread was basically an argument not worth reading!

So much for discussing the topic!

colin
Lemon Slice
Posts: 663
Joined: December 10th, 2016, 7:16 pm
Has thanked: 24 times
Been thanked: 114 times

Re: Lyxor

#140301

Postby colin » May 21st, 2018, 11:26 am

Geoff 100 wrote
On Brexit day, FTSE 100 spreads were kept lower than the theoretical value because market makers were able to make money from the momentum. There is a lot of complexity in the details here.


just about says it all really and confirms my point that deterministic equations have no predictive value because in the real world people adapt to changing opportunities.
Perhaps Geoff100 should read his own posts more carefully as he does not seem to understand the argument, or the topic of the thread.

paulnumbers
Lemon Slice
Posts: 442
Joined: November 4th, 2016, 2:15 am
Has thanked: 32 times
Been thanked: 111 times

Re: Lyxor

#144823

Postby paulnumbers » June 10th, 2018, 4:06 pm

hiriskpaul wrote:By the way, Some of HSBC's tracker OEICs are very competitively priced now and we'll worth a look. For example the C class European index fund, which excludes UK, has a management charge of only 0.07% and no bid/ask spread (supposedly). You would of course need to be wary of any percentage platform charges with OEICS.


Do you think one might be right to be wary of funds that are single priced? Sure, it may seem like there is no spread, but the work has to be done by HSBC to deal with the incoming/outgoing money, and if they're single priced then the cost falls on all investors rather than the investor doing the buying/selling.

Perhaps single pricing is to your advantage if you want to own for a day, but not if you plan to own for 10 years.

GeoffF100
Lemon Quarter
Posts: 4724
Joined: November 14th, 2016, 7:33 pm
Has thanked: 178 times
Been thanked: 1363 times

Re: Lyxor

#144901

Postby GeoffF100 » June 11th, 2018, 7:52 am

paulnumbers wrote:Do you think one might be right to be wary of funds that are single priced? Sure, it may seem like there is no spread, but the work has to be done by HSBC to deal with the incoming/outgoing money, and if they're single priced then the cost falls on all investors rather than the investor doing the buying/selling.

Perhaps single pricing is to your advantage if you want to own for a day, but not if you plan to own for 10 years.

Vanguard has got rid of its entry fees, e.g. for the global bond fund. Instead, they said that they would increase the price if there was an excess of buyers, and reduce it if there was an excess of sellers ("swing pricing"). Nonetheless, my trades have been at the stated NAV. Perhaps Vanguard would say that the costs are being paid out of the AMC, but if they did not do that they could reduce the AMC.


Return to “Passive Investing”

Who is online

Users browsing this forum: No registered users and 9 guests