The Unique Way CPF Computes Interest: Find out when is the best time to make a CPF Investment

Back in 2018, I learned that CPF computes its interest using a method different from that of typical bank accounts, and shared this information as a footnote to a post then. In a nutshell:

• Bank account computes interest on a daily basis, and then pays you interest monthly (i.e. monthly compounding).
• CPF computes interest on a monthly basis, and then pays you interest annual (i.e. annual compounding).
  
• For the exact same amount, you will get slightly more interest when using monthly compounding as compared to annual compounding.

The CPF Board also explains this in layperson-friendly language on its website:

Source: CPF website, retrieved 28 June 2022.

Based on this, I have been under the impression that the monthly interest is computed in this manner:

A post I came across at Fatty's Finance also interprets likewise:  

via Fatty's Finance

Interest is calculated on the "lowest amount in the month" - sounds correct, and matches with the wording from the CPF website, at least from a layperson's reading of the latter.

After doing my own checks, I believe that this interpretation, as well as my diagram above, is not precisely correct.

Unpacking DBS research analysts' bullish view on SIA

(Initially, I had planned to spend some time jotting down some thoughts about how cooling measures, specifically ABSD, might have the unintended consequence of actually increasing property prices, but I shall KIV that for a later post.)

Given the assessments I had made about SIA in prior posts here, I had made a mental note to keep an eye on SIA's FY21/22 financial results, which are due to be announced tomorrow, 18 May 2022. Late this evening, I came across Business Times coverage (link, paywalled) of an analyst report (pdf link) by DBS research, giving a BUY call on the airline. I was intrigued for two reasons:

1. Timing: By no means I track research reports regularly. However, it is uncommon for research reports to be published by analysts one day before a company's FY earnings announcement. Typically, analysts try to take into consideration the latest guidance/outlook as well as financial figures, so it struck me as very odd that the DBS equity research team decided to squeeze out its company update for SIA a day prior to the airline's earnings announcement. 
   
   
2. Thesis: Readers of my previous posts would be aware of my view that the MCBs loom over SIA minority shareholders and present a significant dilution threat. In addition to being extremely bullish on SIA, DBS research also states that they "treat the MCBs as debt instead of equity, as [they] see SIA redeeming the MCBs within 10 years, and deduct the accrued interest at end-FY23/24F." This runs entirely counter to my views and made me wonder if I was missing something. 

I felt a need to jot down my thoughts in response. Here goes:

Strange Risk Assessment by DBS Analysts

The analyst report is exceptionally bullish - with "street-high earnings estimates" and valuation based on a P/BV that's 1.5 std devs above the stock's 10-year mean P/BV. I have no qualms with these, since that is their view. But look at the following paragraph (extracted from p1):

Huh, how can this be a risk? Are the DBS analysts saying that the risk to their (bullish) view is that they might not be bullish enough? Cos that's what the paragraph says. Pretty darn weird. One would expect an assessment of risks to be something like: "recession may dampen travel demand" or "new Covid variants may lead to dialling back of reopening" or "airline runs into manpower issues and unable to hire and re-train quickly enough", but no, we have the above paragraph that presents a strange, almost absurd assessment of the risks of the report team's stated view. 

Unclear how redemption of MCBs can be funded

Granted, any redemption is likely to occur closer to the conversion date of the MCBs, i.e. around 2030/2031, and this is beyond the time-frame of the report. Still, based on the analysts' FY2022/23 and FY2023/24 forecasts (net profit estimated to be $500 mil and $1 bil respectively), I believe that SIA won't be able to generate the $10 bil necessary to redeem the MCBs early, at least not from accumulated earnings between now and 2030/2031.

SIA minority shareholders should be aware that SIA is carrying the $9.7 bil of MCBs on its balance sheet as equity rather than as debt - see 1H-FY21/22 update, p4 of 37. Key implications of doing so:

• By accounting for the MCBs as equity, SIA presents more favourable ratios and financials, such as lower Debt/Assets, Debt/Equity, etc. This is likely advantageous for SIA when it seeks to raise capital, e.g. taking on more debt.  
• By accounting for the MCBs as equity, SIA ought to present key statistics, such as EPS and NAV per share, on basic as well as adjusted basis. [adjusted = assuming MCBs ultimately get converted to shares] Interestingly, SIA did not do so for the 1H-FY21/22 update (see p32) but subsequently reported both basic and adjusted figures in the 3Q-FY21/22 update (see p8 of 10). NAV per share at end-Dec 2021 was $7.45, while just $3.35 on the adjusted basis. The latter makes clear the immense dilutive effect on NAV if the MCBs were held till their mandatory conversion date. 

Notably, DBS analysts treat the MCBs as debt instead of equity, in line with their assumption that the MCBs will be redeemed rather than converted. SIA minority shareholders should pay close attention to valuations and forecasts. From what I can tell (see balance sheet on p8 of analyst report), the DBS analysts adopt SIA's accounting of the MCBs, i.e. carried on balance sheet as equity, even though they say that they treat the MCBs as debt instead of equity. 

I'll give the analysts benefit of doubt that this wasn't an oversight, after all, the redemption (if it happens) is almost certainly to be beyond 2024F. Shareholders should just bear in mind that, regardless whether carried as equity or debt, if the MCBs are to be redeemed prior to the mandatory conversion date, the cardinal accounting equation [Assets = Debt + Shareholders' Equity] still applies. So to wipe the $9.7b of Equity off the balance sheet, SIA will either need to wipe $9.7b off the Assets side of the equation, or pile on $9.7b more in Debt (or a mix of both). The DBS analysts have not shared any views as to how the MCBs will be redeemed, and it is my view that SIA will face a Herculean task in finding a lender willing to extend close to $10b of Debt to the airline for this purpose.

Dividends?? Really?!

The DBS analysts assume that SIA will resume paying a dividend as early as 2023F, with 5.06 cents DPS in 2023F and 10.8 cents DPS in 2024F. (see Forecasts and Valuation table on p1 of analyst report. Separately, note that on p8, the $150 mil expenditure in 2024F for dividends paid would refer to the 5.06 cents DPS paid on ~3 billion issued shares for the prior period 2023F)

I cannot agree with this assessment that dividend payouts will resume, since I believe it would be in SIA's interest to conserve cash if it intends to redeem the MCBs early. Resuming dividends payout simply does not gel with the analysts' view that SIA will redeem the MCBs early. 

===

Overall, I find the DBS research report almost laughably bullish. Of course, DBS and SIA are both under the big T umbrella and in fact have the same Chairman of their respective Boards, so perhaps there's much that I don't know. In any case, it seems that the DBS analysts are more bearish than me for the current FY, estimating an operating loss of $516 mil, as compared to my own latest revised estimate of operating loss between $350-$400 mil. We'll find out tomorrow when SIA releases its full-year results.

In the meantime, I can only hope that SIA minority shareholders are paying close attention to the actual numbers, and not just let themselves be buoyed by bullish media coverage.

===

UPDATE: 19/5/2022

Singapore Airlines released their full-year FY2021/22 earnings yesterday. Actual net loss of $962 mil compared to DBS analyst estimate of $811 mil (loss). Operating loss of $610 mil was greater than DBS analyst estimates of $516 mil as well as my own revised estimates above.

Not surprisingly, the stock price today gave up pretty much all of the boost that the bullish DBS report provided just a day before, as seen annotated below. The timing (and slant) of the analyst report is questionable.

Resumption of travel: What beckons for Singapore Airlines?

[Disclosure: As at time of writing, I do not currently hold any position in SIA, and have no intention to establish a position in the same. I do, however, hold a position in SIA Engineering Company (SIAEC), which is a majority-owned subsidiary of SIA.]

With Singapore announcing reopening of borders with minimal restrictions for vaccinated travellers starting 31 Mar 2022, things are finally looking meaningfully better for travel, including air travel and SIA.

An article in ST on 1 April 2022 discussing SIA, SIAEC and SATS [link; subscription required] compelled me to dust off my thinking cap/analytical lenses, to revisit my previous views regarding holding SIA stock. I last wrote about SIA in May 2021 [link], building on two earlier posts from May 2020 [1] [2].

Rocky Start to 2022

Although 2021 was set to look like a continuation of the positive momentum of 2020 that was driven largely by tech stocks, volatility that emerged around December 2021 and continued through January 2022 has eroded a significant amount of those gains.

I am reminded of my end-2018 post, where I reflected on that year and remarked about how I had ended up pretty much where I had begun. A few years on, with a few clicks on Interactive Broker's fantastic PortfolioAnalyst tool, I was able to generate a couple of reports that contained interesting and informative findings on my holdings under IB.*See footnote.

The risk measures section in the report is something I've paid scant attention to. The entire notion didn't even exist back when I was still using the platforms offered by other brokers in Singapore, because those platforms simply didn't let investors generate any sort of meaningful reports beyond the typical monthly account statement. I'm guessing there will probably be some geek who is still with one of those brokers and has taken it upon him/herself to track their investments separately in some spreadsheet in order to compute these statistics, but I neither have the time nor energy to do so. (In any case, IB offers lower fees than the majority of its competitors, and the competitors' platforms are crappier. So it doesn't really make sense to stick with those competitors, although it appears that a lot of people do, whether due to inertia or simple irrationality, I do not know. Disclaimer: I do have SGX holdings with SCB online trading, so I am occasionally compelled to endure using their platform, but I've been actively seeking alternatives.)

Anyway I digress. IB provided me the risk analysis for 2018:
(to decode: VT is Vanguard Total World; EWS is iShares MSCI Singapore ETF, which I use as an inaccurate proxy for STI; SPX is the S&P 500; and Robust is the alias of my portfolio with IB.)

Risk Measures for 2018

A few more clicks and I got the comparable data for 2020:

Risk Measures for 2020

And finally, I generated the same metrics for the time period since I setup my IB account and started using it, until the most recently available date (3 Feb 2022):

Risk Measures since Inception till Feb 2022

Some observations:

• In 2018, Robust's Max Drawdown was 23.04%. This is almost comparable to the ongoing  decline associated with the recent market volatility: we're looking at Max Drawdown of 22.45% from Robust's peak in early Nov 2021 until its recent trough on 27 Jan 2022. And there's no indication that we are at the market bottom, so there may be more pain in the coming weeks or months. 
• Consider this: Robust's Max Drawdown in 2020 was 30.12% - this was at the onset of Covid outbreak all around the world. If the current bearish market sentiment can't be shaken off, then Robust might approach a drawdown magnitude close to that of the onset of Covid - certainly a worrying thought, but more importantly it's a useful statistic that helps to put things in perspective.
• In terms of Sharpe and Sortino ratios, Robust has outperformed both VT and SPX indices. (Although the EWS index is shown, it isn't a fair comparison since I believe EWS distributes rather than accumulates dividends, and so would be undervalued in all the charts and measures. #See footnote) 2020 was a stellar year with Sharpe ratio above 1, although overall the Sharpe Ratio for Robust since inception remains below 1, at 0.88. I hope to improve this over time. [Sidenote: In calculating downside deviation and Sortino Ratios, IB uses the historical annual return of the S&P 500 since inception, including dividends. This is why the Sortino Ratio for SPX is close to 1 for the Aug 2017-Feb 2022 period.]
• Mean Return for Robust for the full time period is 0.08% daily. This means that if I had $1 million (I don't) in Robust, it would theoretically have generated an average return of $800 every day for the time period. This is a thought experiment more than anything, since there are periods where the portfolio as a whole has declined by 20% or more (see first and second bullet points) but its quite a provoking finding nonetheless.
• Comparing the drawdowns and subsequent recoveries in 2020, the tech-heavy weightage of Robust is evident from its relatively quick recovery of 57 days, compared to the slower recovers of the more diversified indices of SPX (106 days) and VT (110 days). This was because tech stocks benefited immensely from remote working and other Covid restrictions. For the current drawdown, I do not expect Robust to outpace SPX and/or VT in recovery, because the tables have been turned and it is precisely the sector to which Robust is overweight (tech) that has led the recent declines.

*It is truly a travesty that the incumbent brokers in Singapore offer investors no way to generate such reports. This is due to one of two main reasons: 1) the use of individual CDP accounts means that a broker cannot have an accurate picture of its client's holdings at any particular point in time - and by extension, over any particular time period - and hence cannot track returns nor risk measures of a client's portfolio; 2) there is generally a lack of cutting-edge features in investment platforms offered by the incumbents in SG, and most appear comfortable to hobble along with idiosyncratic, feature-thin, and perplexing client-facing products.

# Some info is available on the iShares website https://www.ishares.com/us/products/239678/ishares-msci-singapore-capped-etf#chartDialog that allows for comparison. Using the Aug 2017 to Feb 2022 time period, $10,000 in EWS, with distributions reinvested, would have ended up at $10,492.70. Strangely, the IB captures the EWS return for the same time period as 5.04%, i.e. $10,000 would have ended up at $10,504. Perhaps I'm mistaken, and EWS data reflected in IB reports assumes that distributions are reinvested rather than paid out. But the tl;dr remains the same, performance of Singapore-listed equities has been rather paltry in the last 5 years. Using the iShares site and extending the time horizon to 10 years, the annualised returns for EWS works out to be around 2.3% - notably, this is lousier than returns offered under the CPF Ordinary Account!

A Reflection

 As I approach a significant milestone in my life - the first time in more than 7 years where I will not be getting a salaried income - I thought it would be good to reflect, with the benefit of hindsight, on how I would have approached a couple of key financial decisions.

As with all posts, this is not meant to be treated as financial advice. Instead, some readers may find these perspectives helpful in approaching their own financial decisions.

#1 - Early CPF monies 

Even before I went to university, I had a small amount in my CPF account because I worked in a travel firm during the couple of months between completion of NS and the start of the freshman fall term. This would have been mainly in my CPF Ordinary Account (OA) because CPF allocation rates for young people are set up that way. 

On hindsight, I would have been better off if I have moved most of my OA monies to my Special Account (SA) back then. This would have let me tap on SA's 4+1% interest, which is higher than the OA's 2.5%. The drawback of such OA->SA transfers is that it is one-way, the funds will remain in the SA until I turn 55. This means that, upon such a transfer, I lose the flexibility to use this amount of OA for other purposes (typically housing) in the future. But the amount was pretty small, and honestly wouldn't have made much of a difference by the time I actually got around to a situation where I could potentially use my OA for housing. In that sense, leaving this early amount in the OA rather than transferring to the SA was foregoing potential interest.

To illustrate, assume $1000 left the OA, versus $1000 transferred to the SA. After 10 years, the total amounts including interest accrued would be around $1280 and $1629 respectively. This means that someone who did the OA->SA transfer is 27% better off after 10 years than someone else who did nothing and let the money remain in their OA. 

#2 - Voluntary Contribution to Medisave Account

Almost every Singaporean starting their first job in Singapore will likely experience the joy of receiving their first paycheck, only to have it slightly dampened by the realisation that a significant portion (20%) of their wage must be channeled to their CPF account. While this reduces the amount of disposable income, it actually presents an excellent opportunity to make savvy moves to maximise the longer-term returns from CPF, and the key is to start early.

I recall focusing mostly on the non-CPF part of my salary, i.e. the "cash", and being quite fixated on maximising the interest I could earn from high-yield savings accounts such as OCBC360 and later DBS Multiplier. There was a point in time when OCBC360 was paying 3.05% interest on balances of up to $50k or so, and because of all the hype and marketing, I was more focused on this and largely left my CPF monies as-is. On hindsight, I could have optimised my CPF returns by being more deliberate.

The CPF scheme is fairly complicated, and I don't think I want to cover too much of that here. There's a lot of material available online such as this DollarsAndSense post that readers may find helpful. I wasn't aware back then, but on hindsight I would have paid more attention to the amount in my Medisave Account (MA) and considered a voluntary top-up to this account (sometimes referred to as VCMA). The MA earns the same interest as the SA, i.e. 4% plus an extra 1% on the first $60k of combined OA/SA/MA balances.

In addition to providing income tax relief, the key benefit of topping up the MA is to be able to reach the MA ceiling, known as the Basic Healthcare Sum (BHS), as early as possible. The BHS stands at $63,000 as of 2021 and increases yearly. Upon attaining the BHS, any subsequent CPF contributions that are supposed to go into the MA would flow into the SA instead (assuming one's SA is below the prevailing Full Retirement Sum (FRS) which is almost certainly the case for anyone in the first few years of work). May help to think of it as follows - once the MA "bucket" is full, any overflow from the MA bucket ends up inside the SA bucket. With both the SA and the MA allocation now flowing into the SA, the speed at which the SA grows over time is accelerated, putting the individual on a quicker path to achieving the FRS.

#3 - Retirement Sum Topping Up Scheme or OA-to-SA transfer

The other voluntary contribution covered in the DollarsAndSense article is the Retirement Sum Topping Up Scheme, sometimes referred to as RSTU. If given the option, I would personally have done the VCMA first, then considered the RSTU, but to be frank either option is okay. Similar to VCMA, the RSTU also provides income tax relief, and involves putting cash into the CPF SA to benefit from the 4+1% interest. As I mentioned earlier, once the cash goes in the SA, it will remain there until age 55, so there are some drawbacks compared to having the flexibility of cash. But as the higher-yield savings accounts began to impose more stringent requirements and/or lower overall interest rates, I ought to have considered either VCMA or RSTU as an alternative to put away cash that I did not have a near-term need for. This would have been especially applicable to any cash that was not earning the higher interest rates (banks only gave higher interest on the first $50k-$70k, and a paltry interest beyond that). 

Another way to earn tap on the higher SA interest rate would be to make an OA-to-SA transfer, which I did. At that time, options to invest CPF monies were limited (Endowus - see this post - wasn't an option yet) and I wasn't keen to incur the various recurring fees charged by CPF Investment Scheme (CPFIS) agent banks. And as shared earlier, allocation rates for young people skew heavily towards the OA, so I ended up with quite a lot of money in my OA and not a lot in my SA. In order to tap on the higher interest rate, I moved a portion of my OA to the SA (one-way transfer). On hindsight, I ought to have done this earlier, especially because for the extra 1% on the first $60k of combined OA/SA/MA balances, only up to $20k of the OA balance is eligible - see useful infographic from CPF. This means that I should have gunned for $40k in my combined SA/MA as early as possible. To illustrate more clearly, using three hypothetical persons all with $60k total CPF balance:

• Person A: $45k in OA, $15k in SA+MA. Person A will earn 2.5+1% on first $20k of the $45k, in OA and 2.5% on the remaining $25k in OA, and 4+1% on the $15k. This is roughly $2075 per year in interest earned, or a blended interest rate of 3.458%.
• Person B: $20k in OA, $40k in SA+MA. Person A will earn 2.5+1% on the $20k and 4+1% on the $40k. This is roughly $2700 per year in interest earned, or a blended interest rate of 4.5%.
• Person C: $10k in OA, $50k in SA+MA. Person A will earn 2.5+1% on the $10k and 4+1% on the $50k. This is roughly $2850 per year in interest earned, or a blended interest rate of 4.75%.

I was sort of like person A for quite a while, before finally moving some OA to SA monies and becoming more like persons B/C. Later on, Endowus came into the picture, so now instead of transferring my OA to my SA, I would invest my OA into equities and let that grow over time. Even so, the OA-to-SA transfer remains lucrative because it offers 4% interest risk-free. Of course, if readers foresee needing to tap on OA for housing purchase, then they should plan accordingly, but otherwise as shared in #1, there is  sizeable interest foregone if CPF monies are left untouched in the OA for prolonged periods.

#4 - Set up SRS account

Preamble: the Supplementary Retirement Scheme (SRS) and CPF are two completely different things. If you are not clear, I suggest that you read up elsewhere so that you are clear and do not confuse the two. 

After working for a few years and building up an adequate emergency fund, it may be worth contributing to SRS to enjoy income tax relief. The catch? SRS funds can only be withdrawn when you reach the retirement age at the point that the SRS account is setup. So there's no reason not to set up the SRS first, put in $1, and lock in at the current retirement age (62 as of 2021). If you wait, then you lose the flexibility because eventually your SRS funds can only be withdrawn at an older retirement age. So just set up the SRS account with your preferred local bank as soon as possible.

The question of when to contribute to the SRS account (to enjoy income tax relief) is a more complex one, as it involves individual circumstances, i.e. one's income tax bracket and what one plans to do with the tax savings, so I won't cover it here

Triggering of the Additional MCBs

What does this mean for Singapore Airlines, which has just experienced its "toughest year in its history"?

I'm going light on the narration, because I think the numbers really speak for themselves. I've also shared my thoughts back last year in May 2020 [1] [2] about the 2020 rights issue.

Singapore Airlines pre-covid market cap in July 2019, when travel market was booming: $11.5 bil

Just before the 2020 rights issue, market cap of around $7 bil, on 1.185 billion shares outstanding.

After 2020 rights issue, increase to 2.963 billion shares outstanding. At a low of $3.20, this meant a market cap of around $9 bil. Based on today's closing share price of $4.70, market cap of around $13.9 bil. (Yes, more than pre-covid! But don't forget, SIA raised $5.3 bil cash from the rights issue alone)

If the 2020 MCBs are converted in 2030: 1.3 billion more shares (total 4.26 bil shares)

If the 2021 MCBs are converted in 2031: 2.3 billion more shares (total 6.56 bil shares)

If we assume in 2031 that SIA market cap reaches $20 bil + no early redemption of either tranches of MCB, we are looking at $3.05 per share.

If, on the other hand, we are think that SIA will be able to redeem the first ($3.5b) and second ($6.2b) tranches of MCB before maturity, we are looking at the airline somehow being able to raise a massive sum of money to redeem these bonds. We're talking in the range of $12.4b (if redeemed at the 5th year) to $16.5b (if redeemed at the 9th year).

The numbers well and truly speak for themselves.

Pondering about ticker symbols

So the other day, I was wondering whether ticker symbols had any impact on stock price performance. Specifically, I was curious about whether tickers that came earlier in the alphabet (e.g. AAPL, AMZN) performed differently as compared to tickers further down the alphabet (e.g. TSLA, ZTS).

Impetus for the Idea

The thought struck me while I was looking at my portfolio via Interactive Brokers on my phone. Because of the limited screen size and the default alphabetical sort order of my positions, I observed that I tended to pay more attention to counters that were displayed upfront (those with tickers earlier in the alphabet) as compared to counters that required scrolling in order to be visible (those with tickers later in the alphabet).

Was it just a coincidence that FAANG stocks - Facebook, Apple, Amazon, Netflix and Google - all have tickers that fall within the first half of the 26-letter alphabet, with the sole exception of Netflix, which just tip overs into the 2nd half, with N at the 14th position? A quick look at the five largest public corporations in the world (by market capitalization) [Wikipedia] also showed that the top five, throughout 2020 as well as at 2021, were all of the early-alphabet category. Intrigued!

[factoid: at time of writing, I held only two positions with tickers in the 2nd half of the alphabet: QQQ and ZTS. The rest of my portfolio comprised tickers in the 1st half of the alphabet, i.e. A to M]

Tangential finding from an earlier study

I briefly googled to see if there was any past study done on this (didn't come across, so perhaps an area for novel research!) but instead I found an excerpt of a 2006 Princeton study that found correlation between popularity of a stock after IPO and the ease of pronouncing a company's name. In addition, the study drew similar conclusions about the company's ticker symbol, stating that "...all else being equal, a stock with the symbol BAL should outperform one with the symbol BDL in the first few days of trading."

Now, isn't that fascinating? 

While the study's findings were different from my initial question in two significant ways - (i) it focused on the ease of pronunciation, rather than the alphabetical order; and (ii) it only looked at stock performance right after IPO, whereas I was interested in longer-term performance - the fact that such an observation could be gleaned from real-world data makes me fairly convinced that there could be some sort of association between a stock ticker's alphabetical order, and perhaps either the stock's performance, or interest in the counter. (Interest not necessarily translating to positive returns.)

Basis for my hypothesis

One of the reasons I think such a relationship could exist is the recent emergence of DIY stock trading. As recently as two decades ago, would-be stock purchasers would typically go through a broker, making a phone call to instruct their (human) broker that they wanted to purchase XX shares in YY company at AA price. The would-be purchaser typically would not view any kind of alphabetically-sorted list, before making their instruction. (This is not to say that the would-be purchaser simply plucked their intention out of thin air; they would, in all likelihood, mull over the names - or perhaps tickers - of various companies, before determining that they would want to purchase YY company rather than say ZZ company. The point here is that the would-be purchasers is unlikely to have encountered an alphabetically-ordered list of any kind in the course of their evaluation.)

Today, the typical mode to conduct such a transaction would be via an online broker, either via a web browser, mobile app, or the like. Such platforms provide the added ability to easily view one's portfolio, e.g. through a dashboard. As with any kind of list, there needs to be a default sort order, and I venture that for almost all such brokerage platforms, the default sort order would be alphabetical ascending, placing tickers like AAPL and AMZN at the top, GOOG a bit further down, TSLA further down still... you get the idea. Humans have a natural tendency to cycle through a list from top to bottom, which leads me to believe that there is a possibly-unresearched question of whether ticker symbols correlate with attention paid to the stock, translating into potential interest, and potentially translating into outsized performance. (The point that potential interest may not translate into actual demand for the stock is not lost here, though in this era of meme stocks, I dare hazard that "interest" in a stock tends to translate into positive movements in the stock price.)

Possible angle for research

Acquiring the data required for analysis is probably not insurmountable, given how stock price information is widely available on the internet. The main challenge would be deciding how to measure the dependent variable, and how to control for the various possible interacting factors.

For instance, if we took, say, all 3000 listed companies in the Russell 3000 index, we would expect to find some heterogeneity in the frequency distribution of the leading alphabet of the ticker, for instance, more tickers starting with 'A' than tickers starting with 'V'. We can control for this by taking the average performance for all tickers starting the alphabet, although we may have few data points for certain starting alphabets. Depending on the exact nature of the question we are trying to answer, we could even aggregate our data (e.g. into two groups: Group 1 comprising tickers starting with alphabets A through M, and Group 2 comprising tickers starting N through Z.)

It would then be a matter of testing:

• Null hypothesis: the stock price returns of Group 1 is same as the stock price returns of Group 2
• Alternative hypothesis: the stock price returns of Group 1 is different from the stock price returns of Group 2

We would need to consider controlling for various factors, such as:

• Market cap - if the pre-existing market cap of Group 1 is different from the pre-existing market cap of Group 2, and market cap has correlation with stock price returns, we could end up wrongly rejecting the null hypothesis (Type 1 Error).
• Industry or sectoral concentration - similar to the above, if Group 1 had a higher concentration in fast-growing sectors (e.g. technology) while Group 2 had a higher proportion of slower-growing sectors, we could again commit a Type 1 Error unless this difference is adequately controlled for.
• Ease of pronunciation - with the 2006 Princeton study finding that ease of pronunciation of company name and ticker symbol was positively correlated with purchases of the new stock following its IPO, we would have to find a way to negate any potential effect of Group 1 tickers being generally easier to pronounce than Group 2 tickers, in order to isolate just the effect of alphabetical ordering.

As you can tell, this analysis would not be a trivial exercise, but I feel that existing methods of quantitative research can be easily applied to study this topic and possible yield some interesting insights.

Closing thoughts - What's in a name?

Having posed this question and then regaled you with my idea of how this could potentially be studied, I'll end off with a slightly more pragmatic thought that circles backs to the Princeton study. Danny Oppenheimer, one of the Princeton co-authors, was quoted as saying that "These findings contribute to the notion that psychology has a great deal to contribute to economic theory."

This statement is effectively irrefutable - psychology surfaces in every aspect of human life, and explains how humans have certain biases, or how our perceptions sometimes depart from reality, etc. We probably still don't know how much influence psychology wields over human behaviour leading to economic consequences. Stemming from the observation that easily-pronounced tickers tend to outperform following the IPO, is there then also a possible relationship between the cognitive ease of associating a company's ticker symbol, with object of the company itself, typically represented cognitively by the company name?

On US exchanges, key companies often have ticker symbols that are easily-recognisable variants of the company name, e.g. Apple/AAPL, Tesla/TSLA, General Motors Company/GM, Advanced Micro Devices/AMD etc. This has evolved to become almost a cultural norm of sorts, and companies considering a listing depart from this unwritten rule at their own peril.

But this isn't a global convention. In Hong Kong, for example, listed companies have a numerical ticker (e.g. 9988 for Alibaba). In Singapore, tickers comprise mix of alphabets and numerical digits. At some level, the leading alphabet of the tickers seems to be derived from the company name (e.g. DBS Group/D05, Wing Tai Holdings/W05, UOL Group/U14), but then there are enough exceptions to make this mnemonic useless - Singapore Airlines/C6L, Hong Leong Finance/S41, to name a few. I put forward the notion that the ticker naming system in Singapore makes it more difficult for laypersons to easily recall company-ticker pairs than the US system. 

Could there actually be economic losses arising from the more-difficult-than-necessary association between company names and ticker symbols? Say DBS Group were listed as DBS instead of D05, Singapore Airlines as SQ instead of C6L, City Developments as CDL rather than C09, and Guocoland as GCL rather than F17. Would this have increased market participation and efficiency by lowering, perhaps just ever so slightly, the mental gymnastics required to figure out what is what? I mean, part of the reason why GameStop shot to meme fame was because people could easily input GME and satiate their curiosity on the latest ongoings, rather than having to guess whether it was G48 or F19 or A83. By removing just one additional hurdle, the human brain is freed up just enough to do one additional task - and that task could very well be making a stock purchase decision.

I'm going to put my money that Grab's upcoming listing via the Altimeter SPAC is almost certainly going to gun for the GRAB ticker symbol. The notion that Grab would prefer GB or G01 or GR413 or some other variant over GRAB is beyond absurd. We should then question whether absurd practices, perhaps due to legacy reasons, continue unabated on our very shores, and at what cost.