...Part 2Share Price Looks Great – how’s the 9-bagger of the past 12 years!For all those making snarky digs about the share price now compared with a high point 10 years ago, I say looking in the rearview mirror is a dumb way of investing.
@Sector scoffed at me 12 months ago on the virtues of MSB vs CYP and said that all he needed to know that he backed the right horse was to look at the historic 1,3 and 5 year share price performance between CYP and MSB – well, he was cherry picking dates with CYP at a high back then - if he’d sold his CYP then he would’ve received a 20% premium to the MSB price at the time – so instead of one CYP share now worth $0.70, he’d have 1.2 MSB shares (current MSB price $A3.50) worth $4.20. That’s a 6-bagger he’s missed in the past year by using historic share price performance as his guide. If CYP heads back down to 60c, it’ll be a 7-bagger of regret.
As I showed a few weeks ago, the all-time VWAP of MSB(going back to listing in 2004) is around $2.80, and the price is currently 25% above this so the bulk of people are ahead, and most people on HC who owned at $9.95 have told us they have since averaged down.
So why do the shorts keep on insisting that the $9.95 high is some sort of yardstick, when the far more relevant measure for the vast bulk of people is the recent $A1.02 in March? The $1.02 price was only 4 months ago, so most of the current shareholding base would’ve owned at that price and it is a far more relevant measure than a price 10 years ago in what was almost a totally different company to what MSB is today. If you want to judge yourself on past prices, celebrate the 3.5x bagger we’re sitting on in 4 months, or compare to the $0.50 price in 2008!
You don’t see the shorts strutting around saying “what a miracle the company is”, the price is up from $0.51 on April 4
th 2008 to a recent high of $4.45 – nearly a 9-bagger in 12 years, OR 20% p.a FOR 12 YEARS! - while the ASX200 is only up from 5600 on April 4
th 2008 to 6024 now – up 7.5% IN TOTAL IN 12 YEARS (also note that CYP started 2008 at $4 and was $2 by mid-April 2008 and is only 70c now = disaster for people holding since then – sure, CYP may have been a different company back then, but so was MSB). You can twist past prices and cherry pick any period in the past that suits your story to give any answer you want – they are irrelevant to whether you should own MSB or any other stock NOW.
More importantly however, do your own analysis to judge whether the current price is below or above your long-term valuation. If below, hold or buy more; if above consider selling. That’s a much more sensible way to invest compared with crowing or crying about past prices.
Have a look at the section on valuation at the end to see why I think MSB is currently worth $A10 to $A20.
I’m not talking about trading here, but I think everyone who has followed the convoluted twists and turns of our resident trader realises he’s like most traders who either underperform or lose the lot when they get tempted into leverage. Traders who claim to make small amounts of money on most trades are deluding themselves – the professional traders aim to get 55% right and 45% wrong, and the best traders in the world might get 60% right. So when you read about a trader getting almost every trade right, it raises suspicion and should be ignored (or investigated).
Mesoblast had a stumble when its global partner Cephalon was taken over by Teva and then Teva decided to renege on Cephalon’s agreement with funding trials and commercialising MSB’s products.
In December 2010, after the Cephalon deal announcement of Cephalon paying $130m upfront, and with Cephalon planning to acquire a 20% stake in the MSB at a 45% premium to Mesoblast’s 30-day share price average, the stock gained 21% on the announcement to a then record high of A$4.05, before moving up to $A9.95 in the following year after investors comprehended the benefits of the partnership.
Maybe that $A4.05 price is a more relevant comparison for the ninnies who want to insult the company as a non-performer as the Cephalon deal was to fund all trials through phase 3 and gave Cephalon worldwide rights in most of MSB’s main product candidates at the time. Now, MSB has funded and completed its own phase 3 trials in its three tier 1 products and will have completed the Covid-19 ARDS trial by September. It is currently trading around the pre-Cephalon announcement price, yet it retains almost all of its global product and the product suite now is much bigger now than back in 2010. The market cap reflects this, as the number of shares now are more than double the issued capital back in 2010.
MSB only just starting to hurt index huggers who don’t own it…There’s a bit of smoldering there for the index huggers who don’t yet own MSB. It only officially entered the ASX200 on June 22, and at a price of $4.14 from the close on the preceding Friday, so it actually slightly helped the index huggers who didn’t own it for the last week of June.
The price since June 30 is up 8.0% (at $A3.51), while the ASX200 Accumulation Index is up only 2.1%, so MSB is already outperforming for the new Financial Year by 5.9% - that’s pretty hefty outperformance, but as MSB is only a bit over 0.1% of the Index, it doesn’t hurt too badly, or much at all (yet). This is the first full month in the new financial year and fund managers will start to see it turn up on their radars when they do performance attribution at the end of July. MSB is currently the 34
th best performer in the ASX200 for the new financial year – well behind a few of the small miners and explorers (ORE, OZL, SLR) up 21% to 37% and NWL (Netwealth) up 33% so far since June 30. Among the big stocks, FMG and QBE are both up 17% since June 30, with Afterpay up nearly 15% the only other largish stock up by more than 10%. MSB is not yet important to big instos – but when the share price is closer to $A10, many of them will be forced to buy or risk unhealthy underperformance and volatility of their portfolio relative to the ASX200 Index.
…But it’s definitely burning the sharts
While the MSB shorts over the past 5 years have lost money, the real pain has been felt by the more recent shorts (the sharts) who have been short since March when the price shot up from $1.02 to the current level of $3.51. The price now is close to the highs of the past 5 years, and the borrowing rate has averaged around 12% over that time, so people shorting for any length of time have been smashed by the cost of borrowing stock – and those sharts who are short near the March lows have been horribly burned.
The dollar value of the short position is at its highest of the past 5 years. There were 36.16m shares short on July 14, and they are burning worse than the Bastille. Shorts appear to have started covered in the past week, however, at Friday’s high of $3.58 and the most recent short reported number of 35.35m shares plus another 6m of MSB equivalent shares shorted via MESO ADRs on the Nasdaq – a total short of nearly $150m! The previous high was 41.6m shares shorted in Feb 2019 (and Nasdaq shorting was probably less than 1m shares, but I don’t have the number) – but the price back then was only $A1.17 – so the previous high dollar value shorted was only $50m that’s one-third of today’s short exposure.
This is a massive bet by the shorts in a stock which is due to report 2 blockbuster phase 2 trials in the next month or so - Heart Failure and Chronic Lower Back Pain – both due “mid-year”. Plus in 3 weeks, we’ll have the Advisory Committee of the US FDA holding a full day hearing into potential approval for paediatric aGvHD. Plus, 30 days after the Covid trial has recruited 90 patients, we could see the trial ending for “overwhelming efficacy” – I previously indicated that would easily be the case if Standard of Care results in 60% or more deaths in moderate/severe ARDS and MSB cells result in very few deaths (both numbers appear to be likely to me). Plus, there’s the likelihood of one or more partnering deals with at least $US100m upfront fees for each announced around the time of any successful trial results.
I think a good deal of the price rise since March has been propelled by some of the sharts covering, and then others putting on new positions. The sharts and some of their successors have burned over $60m in losses since the $1.02 low in March this year – with 23.4m shares short in Australia and just under 1m short on the Nasdaq.
Price leads in Aussie market – due to most volume and most shorts in MSB vs MESOI have previously shown quant work showing that the Mesoblast price is set where the most volume trades – and well over 90% of the time that is in Australia. Even the large global instos tend to hold the bulk of their Mesoblast position through the MSB Australian listed stock rather than the US Nasdaq listed MESO ADRs (you can easily see that in M&G and Capital’s substantial shareholder notices).
I think you can see the pain the sharts are feeling when good news comes out such as the announcement of the August date for the Advisory Committee for aGvHD, causing a share price rise of over 10.4% on more than 8m shares traded in Australia on Tuesday this week - the day of the announcement. That’s why didn’t we see such a big rise in the US market that night – the short position in the US Nasdaq MESO ADRs is much smaller than the short position in MSB on the Aussie market. Furthermore, many of the Americans who were short the ADRs, and desperate to cover when the news was released, could’ve bought MSB stock immediately in Aussie trading time to hedge their position in the US.
To try to make back some of these horrible losses, the shorts generally stand back on days of announcements and let the price rise until it starts to peak out – when they see the last buyer entering and the price start to waver, they jump in to try to attack it again and try to force the price down to make a few profits to offset their losses. They can only do that while no new net buying comes in – once the longs sense that the short selling is close to finished , then the longs move back in and their buying will force the price back up and while the shorts may try to resist this for a while, they will eventually have to raise the white flag – especially if the announcements which are due by end September are positive.
So, despite the fact that I think some shorts have covered since March, others have stepped in to try to make some profits on quick trades and the total net short position has risen to a record high in dollar terms, and close to a record high in number of shares. We are still below the 8.3% high in terms of % of the company sharted back when Capital was first selling out – but I think that had probably been telegraphed, with Capital losing funds under management and then losing its long-term fund manager who had supported the holding in MSB.
I have shown in the past that most of the liquidity is in the Aussie market, so that’s where the price is set. There have been a couple of times when US liquidity has gone up dramatically (late April was a great example when 30m ADRs traded (150m shares) in a day – that sort of liquidity will obviously blow Australian trading volumes out of the water – so if the Americans really decide they want to buy MSB, they may then start setting the price, and their buying will inevitably spill over into the Aussie market. However, apart from the odd day or two where US volumes have really picked up, the US Nasdaq price move has zero correlation with the Aussie price move the next day.
US insto holdings in MESO and migration of stock from Aussie shares to MESO ADRsAs a side note, liquidity explains why US institutional ownership of the MESO ADRs appears so low – the big guys hold the bulk of their holdings in the local Aussie traded MSB stock so that they can trade it quickly if they need to, and because hidden costs of trading (buy/sell spread etc) have been lower in the Aussie market. Quoting institutional ownership by just looking at the MESO ADR holdings is meaningless if those instos hold their bulk of their shares on the Aussie market.
That has been slowly changing this year – with the number of ADRs fairly stable since the company sponsored ADRs first listed in 2015 (8.535m listed in November 2015). The number of ADRs jumped to 10m in April this year and have since risen to 14.2m. I suspect this will continue as more US instos decide to buy MSB after possible positive trial readouts (especially Covid 19) – rather than driving the price up on the Nasdaq market and not getting enough volume, they can buy plenty of MSB shares in the Aussie market and then lodge the MSB shares with the ADR Custodian and have ADR certificates issued. MSB has been trading at a discount to MESO most days recently, so I suspect we’ll see that 14.2m ADRs continue to creep higher.
The rise in the number of ADRs this year is equivalent to 28m Aussie MSB shares being bought and converted into ADRs – if that sort of trend continues, there could well be higher liquidity in MESO than in MSB in the future and the price may actually be set in the US market – but I suspect that is still some years away, and may depend on many factors including potential takeover bids, FIRB and Australian government biosecurity considerations, potential pharma partners wanting to hold shares as ADRs etc.
Apart from a brief moment in 2017, It has only been this year that the MESO ADR price has traded above the price prevailing before the disastrous Nasdaq listing in November 2015. Over that same period, the US Nasdaq Composite Index has more than doubled and the Nasdaq Biotech index has risen 35%. While US biotechs have been an underperformer in the past 5 years, they were up more than 4x in the 5 years prior to that (ie 2010-2015). It’s fair to say that MESO (and MSB) have significantly underperformed the Nasdaq and the Biotech Index in the past 10 years, and it’s not hard to understand that it hasn’t been on the radar of US fund managers. They want performance – and it looks like it is finally starting to deliver!
The red line below shows the price of the US ADRs (broker sponsored ADRs) just before the company sponsored ADRs were listed in 2015. It is only since late-April that the MESO ADRs have traded any significant volume above the pre-company sponsored November 2015 ADR price. Interestingly, it was in late-April 2020 that the huge US ADR trade of 30m MESO went through – indicating a big pick up in interest in the ADRs after 5 long years:
View attachment 2332901MSB is starting to run with other global growth stocks – but plenty to goMSB is starting to be recognised (by me anyway!) as a global growth stock. These are the best performing stocks since the start of 2017, and MSB has a lot of catching up to do. It probably means MSB can trade on a much higher P/E ratio than people are assuming if global financial market conditions continue up until the time that MSB starts making early product sales (in October this year hopefully).
Unparalleled levels of monetary stimulus over the last decade has disproportionately benefited Growth stocks. COVID-19 and fears over economic growth have recently supercharged Growth's dominance over Value. The Russell Growth index hasn't outperformed its Value counterpart by this much since the tech bubble of 1999.
Over the last few years, we’ve seen an unprecedented divergence between Growth and Value as investors have seemingly left the latter to die while piling into a handful of high growth stocks. Europe has just announced a massive new borrowing and stimulus package which should give this trend another kick along, and the US probably hasn’t finished its stimulus measures either.
This should really power MSB ahead once it starts to sell product and should ensure a P/E well above my relatively conservative forecasts in my valuations. I have previously noted that with a prospective gross margin of 60% and perhaps as high as 80% once the 3-D bioreactors and serum free media investments are made, MSB could be a “super growth stock” – earning a Return on Equity way above its cost of capital and significantly increasing the value of the company every time it decides to expand.
The following graph show performance of Growth vs Value stocks in the past decade – both the S&P and Russell Indices:
View attachment 2332904Upside Case For aGvHD
My last post was mainly to explain why I had a base case just for aGvHD of only $A2 to $A3. While writing it, I got carried away and started speculating about the upside – which I seriously underestimated because of an error in converting the AUDUSD. I did the currency conversion the right way in the $A2 to $A3 estimate, but inexplicably underestimated the upside by doing the conversion the wrong way. I didn’t pick it up as the number looked reasonable, but it should’ve been much higher – MSB’s upside numbers never look “reasonable” and that’s why the analysts are constantly applying discounts to their valuations and price targets, to try to get the outrageously high numbers down to a more reasonable level so that the analysts won’t be laughed at.
As
@RUWM correctly pointed out I converted the AUDUSD the wrong way in the upside calculation (a rookie error, and I totally apologise):
Wrong: " giving a total net profit from aGvHD in 4 years of around $US350m pa or $A245m pa on a P/E of 25x. That gives a share price target in 4 years of $A10.50..."
Right: As @RUWM said, the sentences should read. " giving a total net profit from aGvHD in 4 years of around $US350m pa or $A500m pa on a P/E of 25x. That gives a share price target in 4 years of $A21.43...".
If you discount that back to today’s values as an NPV, that gives you $A12.25 at a 15% discount rate, or $A10.33 at a 20% discount rate, just for aGvHD.
That massively high upside number,
JUST for aGvHD, prompted me to revisit some of the analysts’ valuations.
Dawson James makes a great comment regarding the difficulty of valuing MSB:
“This is a complex discussion in terms of how does one value a company with both a commercially approved product, multiple partnerships, and 3 x 3 (three products in three pivotal trials, GvHD, Back Pain, and CHF).”
…And that’s before you have a swing at Covid19 ARDS!
Edison
Let’s start with Edison. Their upgrade in Oct 2019 started to recognise the value of aGvHD.
In October 2019, Edison said of their total MSB valuation:
“
We have materially revised our valuation to A$4.1bn or A$7.56 per share (A$7.20 per diluted share) from A$1.72bn or A$4.02 per share. This is primarily attributable to our revised assessment of the aGvHD paediatric and adult opportunities in North America and in Europe, added value for LVAD use in HF and increased price assumptions for Revascor and MPC-06-ID.”
Here’s Edison’s latest “rNPV” ie risk adjusted Net Present Value of the major product candidates.
View attachment 2333033 So, their aGvHD valuation works out at $A2.11 per share BUT, if you take MSB’s guidance of $US700m peak sales (22% higher) you get a value of $A2.57 per share. Then if you get rid of the probability discount, you get at least another 20% higher ie $3.21. And that’s on a NPV basis using what looks to be a discount rate of 21%. The first year’s sales of Ryoncil for the year to June 2021 are only in Edison’s numbers at around $US28m sales ie only around 100 kids treated – so most of the cash flows are way out into the future and being heavily discounted in the NPV by the high discount rate of 21% (my guesstimate).
Note, as I’ve said before, this rough $A2-3 range for the base case of aGvHD is the Net Present Value TODAY. As time moves forward and the low cash flow years when the product is just ramping up pass by, the NPV will rise by close to the amount of the discount rate (ie by 20% or so).
I’ve simulated the Edison numbers and in 12 months, the NPV of the aGvHD cash flows should be a bit over 20% higher, as the first year’s profit is probably in their figures at less than $US10m after probability discounts etc. So, in a year’s time, assuming no probability discount, and assuming MSB’s guidance of $US700m peak sales, Edison’s valuation of aGvHD should rise to $A2.36 bn, or $A3.86 per share.
Finally, let’s drop the onerous discount rate to 15%, and the long-term valuation should rise by around 50% (ie gives aGvHD a value in a year of $A5.80 per share using all of the assumptions above). Dropping the discount rate to 10% would give aGvHD a value of $A8.88 per share, based on my simulation of what Edison is projecting). So, that’s getting closer to my P/E based upside valuation of a bit over $A10 to $A12 per share for aGvHD. The P/E based valuation is still significantly higher because P/Es are currently at record highs so that inflates my valuation, and Edison’s discount rate is overly conservative (IMHO) which pulls their valuation down.
As I said right at the start of this note – valuations for MSB or its underlying products can almost be any number you want to pluck out of the air at the moment. The important thing to know is what factors are the valuations sensitive to and what’s the potential upside and downside.
Now, I don’t have the exact pattern of cashflows that Edison is assuming, not the exact discount rate nor exact probability of success that Edison are assuming here, but I’ve attempted to show how their valuation for aGvHD can increase over the next 12 months as the probability of success rises towards 100% and using a lower risk NPV discount rate. These numbers are rough, but are the best I can do without Edison’s underlying model.
Note 1. There’s still ZERO in there for Covid19 (even though the 12-16 week trial is now 10 weeks through and overwhelming efficacy could result in stopping the trial in a shorter time frame and fast approval could be gained by label extension of Rem-L once it is approved for aGvHD by end September).
Note 1b. There’s also zero for the recently announced related study into children with MIS-C due to Covid-19
Note 2. that there’s still ZERO for Crohn’s Disease (even though the trial has completed and we’re waiting for the readout and possible label extension once Rem-L is approved for aGvHD.
Note 3. The maximum annual sales forecast for Acute Graft vs Host Disease is only $US574m despite MSB publicly estimating $US700m (US, EU, paediatric and adult) the likelihood of Rem-L extension into several other inflammatory diseases (including COPD where trials have been completed; chronic GvHD where Joanne Kurtzberg has an IND; EB study in Japan; HIE study in Japan; etc)
Edison then take this total valuation of the projects, adjust for cash and expenses and derive a total share price valuation of $A7.53 per share:
View attachment 2333036 This July 1
st total valuation number of $US5,204.5m is up from $US4899.7 in March, a rise of 6.2% in 15 weeks. That’s rising at a rate of 22.8% pa, mainly due to first sales and profits being nearly 3 months closer. So, it appears Edison is using a discount rate of somewhere between 20-25%. I’m going to assume it is 21% as that’s close to the rise in the valuation for Heart Failure, Back Pain and “On-Hold Projects”. Graft vs Host Disease went up a little more, but I presume they could have slightly increased their average probability of GvHD success (they use a range of 50-80%). That means they will increase their estimated value of MSB by 21% pa even if there are no new announcements, as long as the time frames for product sales stay the same.
The total Edison valuation increased mainly due to rolling forward their NPV and higher net cash following the US$90m offering in May, which was partially offset by increased expected manufacturing expenses. On a per-share basis, the increase in total valuation was offset by a higher number of shares outstanding.
Here’s the previous valuation of main products on 16 March:
View attachment 2333039Note that the valuation of Ryoncil rose by 7.7%; Heart Failure (Revascor CHF and LVAD) rose by 5.8%; Chronic Lower Back Pain or DDD (MPC-06-ID) rose by 5.7%; “on-hold projects” rose by 5.75%.
If you want to look for upside to Edison’s valuation based on Heart and Chronic Low Back Pain getting FDA approval, we could increase the Probability Of Success on these 2 products to 100% and double their valuation – ie add $A3.765 bn to the valuation. That could add to a $A1.615 bn valuation for aGvHD at 100% probability (assuming the current POS for aGvHD is 80%) and lift the total NPV valuation to $A9.984 bn, or $A16.34. That’s the value TODAY using a 21% discount rate – so in a year, that valuation will be nearly $A20 per share due to the fact that they have very little cash flows assumed in the next 12 months.
So, that’s why I say analyst valuations give a range of $A10-20 per share over the next 12 months BEFORE you include anything for Covid-19 ARDS (or several other label extensions which are possible for Rem-L).
Dawson James
Huge potential upside to $US15 MESO price target with low POS and high discount rate in NPVValuation.
This is a complex discussion in terms of how does one value a company with both a commercially approved product, multiple partnerships, and 3 x 3 (three products in three pivotal trials, GvHD, Back Pain, and CHF). We model each product out to 2030. We provide a detailed explanation of our assumptions (pricing, timing) for each therapeutic model, and then "haircut" our estimates by a probability of success factor, based on the clinical stage of development and our assessment of the indication. For well-established companies with highly predictable revenues, we typically select a risk rate (r) of 10 percent, for early-stage growth companies like Mesoblast, we select our maximum risk rate of 30%. We assume dilution (we never let the projected balance sheet go negative) and use a fully diluted 2030 projected share count. These factors are then applied to our Free Cash Flow to the Firm (FCFF), Discounted EPS (dEPS), and Sum-of-the-Parts (SOP) models, which are equally weighted and rounded to the nearest whole number to derive a $15.00 price target. Our model has previously assumed dilution, and as such, the recent raise does not impact our valuation.
How Does Clinical Success Change the Projected Valuation?
For example, we assume just a 40% success probability in the CHF indications (even though the trial is pivotal). If Mesoblast announces positive clinical data, it suggests the probability goes up. At 100%, this change alone would drive a substantially higher valuation target.
So Dawson James is telling us that their valuation target could be driven substantially higher. The leverage is enormous. The CHF indication's valuation could increase 2.5x if positive clinical data is announced, and the 30% discount rate means their valuation will increase 30% pa for the whole company in the early years before significant cashflows begin.Bottom LineWell, this turned into a long note, but I’m hoping it will be useful to people as it addresses most of the issues which have been raised on HC recently.I have a feeling we will be getting big news on several fronts in the near future – as if last week’s compendium of news wasn’t enough!So armed with this discussion of upside and downside potential, people can do their own analyses based on whatever is announced.I could have continued on with other analysts’ valuations, but I’ve run out of space, and the answers are all the same – massive upside potential if you remove the heavy discounts to valuations, and not much downside risk if aGvHD is approved by the FDA. Longer-term, aGvHD could justify a price over $A10, and maybe as high as $A20 in 4 years on its own once all the markets of US, EU, Adults and Paediatric are approved, plus spin offs in other areas of inflammation. That may all take a few years, but I’m very happy to continue to hold long-term given my exploration of these numbers and the massive extra upside from other product candidates.Could Ryoncil/Remestemcel-L be worth over $A20 per share in 4 years (excluding its value in treating Covid-19)? Sure it could, if everything works out and P/Es then are are up at the record levels of today. I’m certainly not prepared to guarantee any of that, but I hope it shows some of the tremendous potential of Mesoblast.
The more I look at it, the more I like it. This stock takes a lot to get your head around, and I don’t believe many in the professional market have even really tried yet. It’s one thing to do an intellectual valuation exercise, it’s another to believe it. And it's another again to actually put those enormous valuations on paper - nobody has done that yet (even though Dawson James allude to a "substantially higher valuation target" under certain circumstances.Every time I repeat this exercise, I find new insights and I am moving from thinking a $A20 price tag might be achievable in my dreams, to thinking it is actually looking likely if things stay on track and MSB keep being able to add to their credibility.What about $A200 if Covid-19 ARDS works? Well, nobody knows how to value that, and I don't blame them. The amazing thing is that it could happen very quickly and will be worth a lot of money (whether it's over $A100 per share or not, who knows?) - but a LOT of money could come MSB's way SOON - it's an amazing free option - you don't need leverage in this stock, it's got insane amounts of natural leverage!I think the comments of the Principal Investigators in the past week or so show that they seriously think MSB could hold the key to solving the health crisis in Covid-19 ARDS and that would be a really stunning outcome, even though we are hoping it will happen.The US MESO share price has spent the past three months consolidating the huge run at the end of April. I think the MESO chart is poised right at a decision making point (see graph below), from everything we've been hearing recently things are going very well and with lots more announcements due in the short term, I reckon that US investors will decide to buy and push the price decisively up through the upper red line. Announcements, trial results, partnering deals, takeovers, insto buying, short covering, FDA approvals, label extensions - all due in the short term. Talk about excitement!
NB the MESO ADR price graph below is on a log scale to be able to show percentage moves:View attachment 2333030
(20min delay)