Buying/Selling activity in ETFs vs underlying instruments

Looking to see if I can calculate the buying vs selling activity in gold miner ETFs vs the same activity in the underlying equities like Sibanye Gold. Basically trying to see if buying activity is mostly passive moves or actual stock-picking. I’m using Thomson Reuters currently. I basically am looking for the value traded each day of the underlying instruments but need to see if there are inflows or outflows and the volume of the trades.

So the below chart shows ETF gold miners including Sibanye.

What I did was used the difference between the closing and opening market value of the ETFs and multiplied that by the ETF’s closing price/opening price.

So you want to see if the individual stocks’ trading volume is correlated with the ETF’s? Is this to identify if investors are as actively buying/selling the ETFs as much as the stocks? Because most of the time, buying and selling ETFs doesn’t trigger a purchase or sale of the underlying securities. That only occurs when new shares are created to meet demand over what the secondary and primary markets can handle. Most ETFs that have been around for a while have enough shares out there for liquidity to be met in the secondary market and certainly the primary market.

It still happens, of course, normally as a result of a large institutional transaction.

No basically I am trying to identify if the individual underlying equity’s share price is moving because of passive buying/selling in ETFs (like in vanecks junior goldminer) or because of actives.

Reason I used Sibanye as an example is that its rallied hard on the gold price increasing as many of the gold etfs its included in have rallied. But only 13% of EBITDA is from gold whereas the remaining is from 3E PGMs. They are down 5.5% but with Rh and Pt rallies there should be incremental buyers.

No, the underlying securities shouldn’t move because of buying/selling of the ETF (neither active nor passive) for reasons mentioned above. That’s why ETFs can trade at a premium or discount to NAV. I’m sure there’s some way to arbitrage it away if the spread (not bid-ask; premium/discount to NAV) is large enough. That would impact the individual stocks, but it seems unlikely to happen very frequently.

actually from what i’ve read the etfs trade pretty close to book. if there is a lot of demand and there is a premium or discount then people often deliver the stocks and the issuer creates the etfs, or they redeem the etfs for the stocks. usually if you want to create or redeem you need to do a large batch, and that will affect the underlying.

for the most part though people just buy and sell the etfs to each other in small batches so there is n oeffect.

Okay, so let’s give an example to maybe clarify what I’m saying.

Let’s say the ETF in question is the vaneck vectors junior gold miner etf (includes sibanye as a constituent) and for the examples sake the etf is equally weighted.

If some of the other constituents have larger share price moves than sibanye and the index is rebalanced then the etf would have to either buy/sell some of sibanye. In this case a rebalancing would move the share price of sibanye not active funds individually buying sibanye (for my purposes I think of an active fund as say an asset manager).

etf typically dont rebalance. most etfs are based on market cap. so if price drops then that is their new allocation. the only time they’d need to rebalance is when they replace a constiuent. remember goal of etf is to track the performance of their benchmark. but their benchmark will move based on price changes (which fixed itself), or changes in the constiuents.

Nicely put Nery. Jack, you’re trying to fit a square peg in a round hole. Even active ETFs (for which the market is tiny) still have very, very little turnover. After all, keeping trading costs to a minimum while maximizing tax efficiency are really the cornerstones of why anyone buys ETFs.

The exception is leverage ETFs, but they don’t belong in this conversation.

I’m not quite sure how we can put it any more simply. Buying and selling ETFs on the open market does not directly affect the underlying securities. Hard stop. Only the creation or redemption (read: liquidating the basket of underlying securities) would move the price of the stocks.

Your concern over rebalance is moot because it just doesn’t happen. What can and does happen is when, as Nery points out, an index adds or kicks out securities and the ETF must follow suite. Here the ETF provider is between a rock and hard place because they’re charged with keeping tracking error to a minimum (even active ETFs have custom benchmarks) while also keep trading to a minimum. But it happens and that’s why when the S&P 500 changes adds a stock it generally sees a short-lived price bump. Though even that seems to have been arbitraged away in recent years.


I do want to clarify something. In my haste, I miscommunication the order of liquidity and when the underlying stocks are affected. It goes like this:

Secondary Market - Where all retail trades are done as well as trades for highly liquid ETFs like SPY. Trading ETFs here has no impact on the underlying stocks.

Market Makers - If an order can’t be filled on the open market, the designated market maker will step in and try to fill the order. This generally gets the job done even for very large transactions. Trading ETFs here has no affect on the price of the underlying stocks.

Primary Market / Authorized Participant - When all else fails and someone needs to buy or sell such a large amount not even the market maker can fill it, the Authorized Participant is contractually obligated to create more shares, generally in lots of 50,000 or maybe 100,000. Even at this stage, the AP does whatever they can to minimize moving the price of the underlying securities, although it does have some impact.

Rebalance moves companies among MidCap, SmallCap indexes https://seekingalpha.com/news/3497790

haha some index actually had some changes this Friday, effecting next week. I know this isn’t what you are looking for though lol. But interesting enough!

I think maybe my explanation was a bit poor.

So I realise the ETF tracking the index does not move the underlying security prices in theory.

But think of this example, let’s say hypothetically a fund decides to increase their gold exposure. So they go out and put $200 million in gold etfs. Those gold etfs will have to take the $200 million and go and buy underlying securities. Which would move the prices of the etfs constituents.

The whole reason I was asking this question was gold etfs are up, but Sibanye Gold has been down -11% in the past week. I am trying to figure out who the incremental seller is

if you are buying the etf, and someone who owns it i selling it, then there is no effect on the underlying price.

if no one is selling it, and there is a decent premium on the etf. then someone will bid up the underliers and bundle it up to create new etfs for you.

i dont know how you can identify who is selling it. but its silly to think that just because an etf who has 50 consituents is up, then your 1 and only constiutent should be up as well. there are a many reasons why your single stock is down. its been up 100% in 1 year. it has a lot of debt. it has a negative net income. and many more.