Kristian P. Talvitie
Executive Vice President & Chief Financial Officer at PTC
Thanks, Jim, and good afternoon, everyone. Before I review our results, I'd like to note that I'll be discussing non-GAAP results and guidance, and ARR references will be in both constant currency and as reported. Turning to Slide 19. In fiscal '22, our ARR grew 16% on a constant currency basis and exceeded guidance. On an organic constant currency basis, excluding Codebeamer, our ARR was $1.61 billion, up 15% year-over-year. As Jim explained, our top line strength in Q4 was broad-based. We're executing well against our strategy, and we're continuing to improve upon our strong market position.
Our SaaS businesses across our digital thread and Velocity groups saw continued solid ARR growth in Q4. On an as-reported basis, we delivered 7% ARR growth, 6% organic, due to the impact of FX headwinds, which were $134 million, substantially higher than the approximately $85 million of FX headwinds that we would have expected assuming Q3 ending exchange rates and at the midpoint of our guidance for Q4. Moving on to cash flow. Despite the FX headwinds, our cash flow results were strong, with Q4 and full year results coming in ahead of our guidance across all metrics. Solid collections performance and slower hiring helped to offset the incremental headwinds that materialized in Q4.
And for the full year, our restructuring cost controls and above planned perpetual license revenue, primarily from Kepware also helped to offset the FX headwinds. When assessing and forecasting our free cash flow, it's also important to remember a few things. FX rates were more favorable in the first half of fiscal '22, and the majority of our collections occur in the first half of our fiscal year. Q4 is our lowest cash flow generation quarter, and free cash flow is primarily a function of ARR rather than revenue. In fiscal '22, FX fluctuations created a headwind of approximately $30 million to our free cash flow results. Q4 revenue of $508 million increased 6% year-over-year, and fiscal '22 revenue was $1.933 billion, up 7%.
As we've discussed previously, revenue is impacted by ASC 606, so we do not believe that revenue growth rates are the best indicator of our underlying business performance, but would rather guide you to ARR as the best metric to understand our top line performance and cash generation. On a constant currency basis, our Q4 revenue was up 12% year-over-year, and our fiscal '22 revenue was up 11%. Before I move on to the balance sheet, I'd like to provide some color on our non-GAAP operating margin as I did last quarter.
We continue to caution, because revenue is impacted by ASC 606, other derivative metrics such as gross margin, operating margin, operating profit and EPS are all impacted as well. Still, it's worth mentioning that we're benefiting from the work that we've done to optimize our cost structure. Our non-GAAP operating margin expanded by approximately 300 basis points to 38% in fiscal '22. Moving to Slide 20. We ended the fourth quarter with cash and cash equivalents of $272 million. Our gross debt was $1.36 billion with an aggregate interest rate of 3.9%. During Q4, we repaid $75 million on our revolving credit facility.
Regarding our share repurchase program, as we've communicated previously, we completed $125 million of repurchases at the front end of this fiscal year. Our long-term goal, assuming our debt-to-EBITDA ratio is below three times, is to return approximately 50% of our free cash flow to shareholders via share repurchases while also taking into consideration the interest rate environment and strategic opportunities. Next, Slide 21 shows our ARR by product group. In the constant currency section on the top half of the slide, we use rates as of September 30, 2021, to calculate ARR for all periods.
You can see on the slide how FX dynamics have resulted in differences between our constant currency ARR and as reported ARR over the past eight quarters. I provide a reminder on the previous call that when we set ARR guidance for fiscal '23, we would be providing that at the September 30, 2022 FX rates and restating history assuming those rates. Let's turn to that now. Here on Slide 22, we show the new reporting categories that Jim took you through with historical ARR results recasted using our FY '23 plan rate for all periods.
We post a set of financial data tables to our IR website that has our financial statements as well as these ARR tables. For comparative purposes, the constant currency historicals are at FY '23 plan rates and should be used when forecasting PTC's constant currency ARR results. This is important to consider in the context of our guidance because we provide ARR guidance on a constant currency basis. If exchange rates fluctuate significantly between the end of Q4 and the end of Q1, that would be reflected in our as-reported ARR.
We believe constant currency is the best way to evaluate the top line performance of our business because it removes FX fluctuations from the analysis, positive or negative. Given the sharp moves in FX that we've seen recently, I thought it would be useful to provide an updated ARR sensitivity rule of thumb here on Slide 23. And as you can see, in addition to the U.S. dollar, we transact in euro, yen and more than 10 other additional currencies. Using our Q4 FX rates, the impact of a $0.10 change in the euro would be approximately $38 million positive or negative and the impact of a JPY10 change would be approximately $7 million, again, positive or negative.
And of course, the estimated dollar impact to ARR is dependent on the size of the ARR base. Turning to Slide 24. Since most of the sell-side forecasts have been built, assuming September 30, 2021 rates, I thought it would be helpful to provide this slide as a reference point before I take you through our actual guidance on Slide 25. Slide 24 illustrates what our guidance would have looked like if calculated using our fiscal '22 plan rates. The year-over-year growth rate for ARR is the same 10% to 14% that you'll see on Slide 25. However, the absolute values for ARR are higher given the FX rates versus the FY '23 plan rates.
For revenue, the actuals we report and the guidance we provide are on an as-reported basis, not on a constant currency basis. For illustrative purposes, this slide shows our fiscal '22 revenue on a constant currency basis, assuming our fiscal '22 plan rates. On this basis, the illustrative year-over-year growth rates for fiscal '23 are about 6% higher than the comparable numbers you'll see on the next slide. The range of revenue growth is lower on Slide 25 because our reported revenue is based on monthly FX movements rather than a point in time set of rates, and we faced some tough comps in fiscal '23, particularly in the first half of the year. With that, I'll take you through our guidance on Slide 25. Today, I'll focus on FY '23 and Q1.
In a couple of weeks when we have our Investor Day, we'll focus more on the medium term. For all ARR guidance amount, we're using our fiscal '23 plan rates, the rates as of September 30, 2022. For fiscal '23, we expect constant currency ARR growth of 10% to 14%, which would make this the sixth consecutive year of double-digit growth. This corresponds to fiscal '23 constant currency ARR of $1.73 billion to $1.79 billion. Churn is a key component of our ARR guidance, and we significantly outperformed our churn guidance in fiscal '22 and ended the year with organic churn at 5.6% using our fiscal '22 plan rates. Our medium-term target, as stated back in 2019 was to get to the 6% level by fiscal '24, and we achieved this earlier than expected.
Going forward, I think there's still room to improve our churn rate, but at the same time, I want to be cognizant of the macro environment, and that's the rationale behind keeping our churn assumption essentially flat for fiscal '23. Next, on cash flows. For fiscal '23, we expect free cash flow of approximately $560 million, up about 35% and adjusted free cash flow of $562 million, up about 20%. This includes an estimated $60 million of headwinds caused by FX impacts on our operations as compared to FY '22 FX rates. Our fiscal '22 restructuring is substantially behind us. So the difference between free cash flow and adjusted free cash flow is small for fiscal '23.
Our capex assumption for fiscal '23 is approximately $20 million, and therefore, relative to free cash flow guidance of $560 million, we're guiding for cash from operations of $580 million. In a few minutes, I'll take you through an illustrative FY '23 free cash flow model in more detail. Moving on to revenue guidance. For fiscal '23, we expect revenue of $1.91 billion to $1.99 billion, which corresponds to a growth rate of negative 1% to positive 3%. ASC 606 makes revenue fairly difficult to predict in the short term for on-premise subscription companies, hence the wide range. And remember, revenue does not reflect cash generation as we typically bill customers annually upfront regardless of contract term lengths. It's worth taking a few minutes to explain this.
So let's go on a short detour from guidance and turn to Slide 26. This slide shows a hypothetical example with 10 contracts and different assumptions for four variables: software type, which can be either on-premise subscription, perpetual support or cloud/SaaS; number two, upfront recognition percentage. To keep this model simple, we'll just assume that 50% of the on-premise subscription total contract value is recognized upfront under ASC 606. To be clear, this varies by company and even by product within each company. But here, we're just trying to keep this simple to illustrate the point. Number three, term length. This example uses term lengths ranging from one to four years and four contract size.
This is really total contract value or TCV. In this example, TCV ranges from $1,000 to $4,000. But note, all of these contracts have the same ARR of $1,000. Moving to Slide 27 for certain contract types, namely SaaS and cloud and on-premise support contracts, revenue recognition is ratable over the term of the contract. For these contracts, term length and contract value do not have impact on revenue recognition and revenue aligns with ARR on an annual basis. You can see this here with $250 being recognized every quarter. On the next slide, Slide 28. Here is an example of a one-year on-premise subscription that renews annually.
Due to ASC 606, half of the contract value is recognized as upfront revenue and the remaining half is recognized ratably over the contract term. Since this is a one-year subscription, revenue aligns with ARR on an annual basis. However, on a quarterly basis, there is a difference because of the upfront revenue recognition was $625 in the first quarter and $125 in the next three quarters. Turning to Slide 29. Here, we're highlighting an example of a four-year on-premise subscription. Again, under ASC 606, half the contract value is recognized as revenue upfront. However, ARR and cash invoicing are both done on an annual basis. In this example, the amounts would be 1,000 per year.
In contrast, as this example shows, we would recognize $2,000 of revenue upfront, and the remaining $2,000 will be recognized ratably over the 16 quarters of the term. Moving to Slide 30. The graph compares ARR and revenue for these 10 hypothetical contracts over a 12-year period. Over the 12 years, this model assumes no growth, no price increases nor churn. There's no changing of term lengths, no migrating from support to on-premise subscription or from on-premise subscription to SaaS. Of course, as you know, we actually have programs in place that actively drive each of these dynamics, but it would be much too complicated to start adding these kinds of dynamics into this super simple example.
The green line, which is ARR, appropriately shows flat business performance, 10 contracts at $1,000 each for $10,000 in ARR each year, 0% growth. Whereas the blue line, which is revenue, shows a lot of volatility from year-to-year. Revenue growth rates vary from negative 33% to positive 69% in any given year. And again, while this example is overly simplistic, hopefully, you can see why we keep saying that you can't really look at revenue to understand the underlying performance of our business, nor is it helpful when trying to understand free cash flow dynamics.
What is important to remember is that over the term of the contract, over the term of each contract, every dollar of ARR becomes $1 of revenue. However, how and when revenue is recognized is dependent on the mix of contracts starting and/or renewing in any given year and can vary significantly from period to period. And you can imagine what happens to margins and EPS under this kind of scenario. Next, on Slide 31 is the conclusion. Due to ASC 606, the revenue trend is noise for companies like PTC that sell on-premise subscriptions.
But the point here is, don't worry because free cash flow is really a function of ARR. And this is why we focus on ARR and free cash flow and believe it to be the best way to assess our fundamental business performance. Now returning from that fun side trip to guidance on Slide 32. For Q1, we're guiding for cash from operations of $170 million, free cash flow of $165 million and adjusted free cash flow of $166 million. There's one additional item worth noting related to the cash flow in Q1 and fiscal '23. In Q1, we expect to make a $10 million foreign tax payment related to changes to the withholding tax policy of a foreign country.
Since we also expect this to be substantially -- all of this to be substantially -- substantially all of this to be refunded before the end of fiscal '23, there is no expected impact to the full year financials and we will not adjust for this. Aside from this item, as you model the quarters of fiscal '23, keep in mind that we expect the quarterly distribution to follow a similar pattern as in fiscal '22 and fiscal '21 for the cash flow metrics with over 60% of our cash flow coming in the first half of the year and Q4 being our lowest cash flow generation quarter. Next, on Slide 33. Let's take a look at an illustrative model that uses the assumptions that Jim described earlier and illustrates what you need to believe to achieve the midpoint of our fiscal '23 constant currency ARR guidance at the midpoint of $1.76 billion.
In this illustrative model, which is for 12% constant currency ARR growth, we've assumed that churn worsens by approximately 100 basis points, and we kept new ACV flat. While new ACV is not exactly the same as bookings due to dynamics such as ramp deals and in-year starts, which we've discussed before, it's a close enough proxy for this exercise. I would point out that while the churn increase looks high, that is, in fact, the math of an approximate 100 basis point increase that we outlined earlier in the discussion. And while the macro environment may increase churn, as you can see on the page, a 100 basis point increase would push absolute churn level -- churn dollars above levels that we have not seen for quite some time.
And we have some tailwinds on the churn front, which include term lengths lengthening. In fact, our expiring base in fiscal '23 is only slightly higher than in fiscal '22 despite more than $200 million of beginning ARR. The price increase we implemented in May should also provide some tailwind throughout the year. And importantly, our product portfolio continues to mature and some of the product segments that have higher churn like IoT and AR have seen steady improvement in churn over the past three years and still have lots of room for improvement to get to renewal rates more in line with, for example, Creo and Windchill.
On the flat new ACV assumption, I would also point out that we have slightly more deferred ARR starting in fiscal '23 than we did in fiscal '22. We will have a full year of new sales from Codebeamer versus the two quarters worth that we had in fiscal '22, and the price increases that went into effect should provide a modest tailwind on the new sales front as well. You can see that to hit the midpoint of $1.76 billion, we need to add $188 million of ARR in fiscal '23. Balancing potential macro uncertainties with the momentum we've been seeing in our business, our forecast and given some of the considerations I just mentioned, we believe our fiscal '23 guidance is prudent.
Turning to Slide 34. Here's an illustrative constant currency ARR model for Q1. You can see our results over the past eight quarters. In the far right column, we've modeled the midpoint of our constant currency Q1 ARR guidance. Because our ARR tends to see some seasonality, the most relevant comparison is Q1 '21 and Q1 '22. The illustrative model indicates that to hit the midpoint of our Q1 '23 guidance of $1.59 billion, we need to add $18 million of ARR on a sequential basis. This is less than the $40 million we added in Q1 of '21 and the $37 million we added in Q1 of fiscal '22.
Again, for the reasons I mentioned just a minute ago, we believe we've set our Q1 '23 constant currency ARR guidance range prudently. Let's move on to Slide 35, which shows an illustrative free cash flow model for fiscal '23. I know that most of you model free cash flow using the indirect method, which uses the P&L and balance sheet as a starting point. Given the complexities related to ASC 606 that we spent some time on earlier on this call, there are inherent challenges in using the indirect method to forecast free cash flow for PTC. The model on this slide is based on what we use internally.
I know that looking at it in this way may be unfamiliar to some of you, so please feel free to reach out if we can help. Starting at the top with ARR. Note that the FY '22 ARR has been -- is at the September 30, 2022 rates and that for fiscal '23, we've used the midpoint of our constant currency ARR guidance. Moving down the model. The primary reason why FY '23 professional services revenue is down is because of unfavorable FX developments we've seen over the past year. In addition, we expect roughly 1/3 of our professional services revenue to transition to DxP over time. Moving to cost of revenue and operating expenses, the FX developments over the past year have the opposite effect here.
Compared to revenue for costs and expenses, the FX moves are beneficial. You can see the powerful leverage on our cash contribution, which is a characteristic outcome of a sticky recurring subscription business model, combined with operational discipline. As you can see, expected restructuring payments are materially lower in fiscal '23 because, as Jim explained earlier, we completed the work related to the operational optimization announced a year ago and the related cash restructuring payments are now substantially behind us.
We've also completed the work to rebalance resources across the company to align with our growth opportunities and we did this without a restructuring charge. Moving on, you'll see that other expense is higher in fiscal '23 compared to fiscal '22, this is primarily due to higher interest rate on our revolver, given the higher interest rate environment. Cash taxes are also higher in this model and reflect both higher taxable income as well as the impact of Section 174. And finally, the other category is also higher in FY '23. The main driver here is higher working capital to support continued growth. FY '22 also includes the impact of FX movements.
All this sums up to expected free cash flow of $560 million, inclusive of the $134 million ARR headwind that materialized and consequent $60-ish million headwind to free cash flow. Jim explained earlier, we expect to deliver approximately $560 million under the most plausible ARR growth scenarios, and I'd like to reiterate that. If the macro situation gets worse and our cash generation is impacted, you can expect us to moderate our spending. On the other hand, if the macro situation improves or the dollar strength reverses, this would be favorable for our cash generation. And in that scenario, we'd likely invest more aggressively in our business.
As we start the year, we believe $560 million is a good target. And that's a good segue to Slide 36. First of all, we are prepared for a storm and expect to be resilient in the face of one. From a top line perspective, we serve industrial product companies. And R&D at those companies tends to be quite resilient, so we have a supportive top line backdrop. We've also transitioned successfully to a subscription business model, and our products are very sticky with our customers. Just as importantly, from a cost and operational perspective, we have done a lot of optimization in 2022 that will serve us well going forward. We've already battened down the hatches.
Nevertheless, as we've said in the past, we do not have a Pollyanna type view when it comes to the macro situation. We have a strong track record of disciplined operational management. And in the event of a meaningful downturn, we're prepared to pull additional spending levers to mitigate the impact on our cash flows. Variable compensation would automatically adjust. And depending on the magnitude of the downturn, we would also curtail plant hires, look at backfilling, attrition, marketing spend, travel spend, etc. So with that, I'd like to wrap it up here and turn it over to the operator to begin Q&A.