Which of SPY, FXAIX, VOO, IVV can get the most return over the long run?

Question

I gathered the following data from Yahoo Finance's Historical Data.

I didn't start from March 3, 2011 because the Yahoo could not provide data for March 2011 for one of the ETF or Fund, so I used June 3, 2011 instead.

However, SPY has a 0.09% Expense Ratio, and VOO and IVV has 0.03% ratio, and FXAIX has a 0.015% expense ratio, but it seems VOO can turn $10,000 into $35585, followed by SPY's $35530, IVV's $35520, and FXAIX's $34889. For some reason it is not in the correct order if estimated by expense ratio.

So I think it is not too important for the difference, unless if it is over a million dollars, in which case the max and make $5000 to $6000 more over 10 years, which is still somewhat substantial.

Between the max and min, VOO's $35585 and FXAIX's $34889, there is a close to $700 difference. So if the investment is $100,000, that means the difference is $7000, which is still quite a lot. It is strange because FXAIX's expense ratio is 0.015%, which is the lowest.

However, for the fund FXAIX, can it automatically re-invest the dividend without incurring tax no matter it is in 401(k) or IRA account? For SPY, VOO, IVV, I think the dividend paid out is subject to tax immediately unless if it is in a 401(k) or IRA account, so if FXAIX can re-invest the dividend without any immediate tax, that can be an advantage.

Or is it just the opposite, that if we can put the money in 401(k) or IRA, we don't have to pay the tax until later, while for FXAIX, I do see some dividend and capital gain data in its fact sheet, so does that mean the fund always has to pay tax immediately no matter we have it in our regular or retirement account?

(We might be able to invest $300 or $1000 into each of them and see how they grow over a few years, both in a regular or retirement accounts, except we might be paying tax for some of them for the dividend without knowing it.)

All data obtained from Yahoo Finance Historical Data

S&P 500 (^GSPC)

2011/6/3 ^GSPC Close and Adj Close: 1300.16 
2021/3/3 ^GSPC Close and Adj Close: 3819.72  

So 2.9378845680531622  if divided to get the ratio

SPY

2011/6/3 SPY Close and Adj Close: 130.42    107.35
2021/3/3 SPY Close and Adj Close: 381.42    

so 2.924551449164239  if divided by close
3.5530507685142063  if divided by adj close

(Adjusted close price adjusted for both dividends and splits.)

FXAIX

2011/6/3 FXAIX Close and Adj Close: 46.17   38.04   
2021/3/3 FXAIX Close and Adj Close: 132.72

so 2.8745938921377516  by close
3.4889589905362777  by adj close

VOO

2011/6/3 VOO Close and Adj Close: 119.30    98.54   
2021/3/3 VOO Close and Adj Close: 350.66

so 2.9393126571668065
and 3.5585549015628173

IVV

2011/6/3 IVV Close and Adj Close: 130.88    107.78  
2021/3/3 IVV Close and Adj Close: 382.84

so 2.9251222493887528
and 3.5520504731861196

Answer 1

All ETFs tracking the same index will have roughly the same performance. There is always a small tracking difference but whether this tracking difference will turn out in your favor or not is unpredictable and over the long run it will even itself out. Therefore comparing ETFs based on past performance is pretty pointless.
Instead focus on the following topics

Cost
Costs are one of the few things impacting your performance that are predictable. An ETF with 1% will always perform worse than one with 0.1% if they are tracking the same index. Prefer lower cost ETFs
Cost is not limited to management expenses. It can also include tax treatment brokerage costs (here in Germany hefty commissions are still the unfortunate standards for ETF plans but most brokers have some "partner ETFs" with reduced commissions).

Handling of dividends
Do you want to have your dividends automatically reinvested in an accumulating ETF or dou you want to have money paid out every year by a distributing fund? The first has the benefit of automatically compounding your investment while the latter allows you to earn a passive income without selling assets. Depending on whether you are in the savings phase or in the consumption phase of your investment, this may be different.
Taxes may also play a role here as you want to prefer reinvesting with tax-free gains if possible.

Replications Method
Most ETFs replicate an index by buying shares (physical method). However, some replicate with a synthetic method by buying swaps. If the bank selling the swap goes down the drain, the fund will take a big hit. This is a factor that is often overlooked but can mean in practice that your supposedly low-risk ETF on the S&P500 actually is a higher risk asset that can lose a big portion of its value with the next Lehmann Brothers going down.

Answer 2

An even better strategy (if you're doing monthly DCA) would be to leverage all three of them for tax loss harvesting.

For example, if you invest $100/month into SPY and then it drops by 10% in month 6, leaving some of your lots in the red, then you can sell those lots and invest the proceeds in VOO. Rinse and repeat with FXIAX. By strategically rotating between the 4-5 top ETFs tracking the index, you can maintain your position while harvesting losses, thereby increasing your net return.