SENATE FINANCE COMMITTEE February 10, 2022 9:03 a.m. 9:03:39 AM CALL TO ORDER Co-Chair Stedman called the Senate Finance Committee meeting to order at 9:03 a.m. MEMBERS PRESENT Senator Click Bishop, Co-Chair Senator Bert Stedman, Co-Chair Senator Lyman Hoffman Senator Donny Olson (via teleconference) Senator Natasha von Imhof Senator Bill Wielechowski Senator David Wilson MEMBERS ABSENT None ALSO PRESENT Alexei Painter, Director, Legislative Finance Division; Conor Bell, Fiscal Analyst, Legislative Finance Division. SUMMARY PRESENTATION: FISCAL SCENARIOS - LEGISLATIVE FINANCE DIVISION ^PRESENTATION: FISCAL SCENARIOS - LEGISLATIVE FINANCE DIVISION 9:05:03 AM ALEXEI PAINTER, DIRECTOR, LEGISLATIVE FINANCE DIVISION, discussed the presentation entitled "Fiscal Modeling: Senate Finance Committee Scenarios" (copy on file). He relayed that he would cover the LFD baseline assumptions, compare those to the committee assumptions, and run various fiscal models. Mr. Painter looked at slide 2, "Outline": ?Review of LFD Modeling Baseline Assumptions ?Comparison of Senate Finance Committee assumptions to LFD Baseline ?Fiscal Models Using Senate Finance Assumptions Mr. Painter noted that the co-chairs had asked for different assumptions to compare to the baseline. He said that several fiscal models would be used to run different assumptions, as well as stress tested. Co-Chair Stedman reminded that there could be other modifications to the modelling of the numbers. 9:06:10 AM Mr. Painter spoke to slide 3, "Review of LFD Modeling Baseline": ?Legislative Finance's fiscal model is designed to show policy makers the longer-term impact of fiscal policy decisions. ?The baseline assumptions are essentially that current budget levels are maintained, adjusted for inflation. Policy changes are then applied against that baseline. ?Our default is to assume that statutory formulas will be followed. Co-Chair Stedman thought slide 3 was routine for the committee but was not routine for the public. He asked Mr. Painter to go over slide 3 more slowly. Mr. Painter reviewed slide 3. He stated that the LFD goal was to show policy makers the long-term (next 10 years) impact of proposed policies. He shared that the modelling started with a policy-neutral baseline, in the case of the presentation using the governors budget submission (which was a policy proposal) and show how it grew with inflation. He said that when scenarios were run, policy changes were applied against the baseline. He relayed that the default assumption was that the statutory formulas would be followed and any deviation from statute would be reflected as a change from the baseline. 9:08:25 AM Mr. Painter referenced slide 4, "Review of Modeling Baseline (cont.)": Revenue Assumptions ?LFD's baseline revenue assumptions are the Department of Revenue's Fall Revenue Forecast. This assumes $71 oil in FY23, following futures market thereafter. DNR oil production forecast projects that Alaska North Slope production will increase from 500.2 thousand barrels per day in FY23 to 586.2 thousand barrels per day in FY31. ?For the Permanent Fund, we use Callan's return assumption of 5.86% total return in FY22 and 6.20% thereafter. 9:09:24 AM Mr. Painter turned to slide 5, "Review of Modeling Baseline (cont.)": Spending Assumptions  ?For agency operations, these scenarios assume the Governor's FY23 budget grows with inflation (2.0%). ?For statewide items, the baseline assumes that all items are funded to their statutory levels beyond FY23. This includes School Debt Reimbursement, the REAA Fund, Community Assistance, oil and gas tax credits. ?For the capital budget, we assume the Governor's FY23 capital budget grows with inflation (2.0%) ?For supplementals we assume $50.0 million per year. This is based on the average amount of supplemental appropriations minus lapsing funds each year. Mr. Painter noted that Callan had indicated the inflation growth number in the first bullet would increase, which would be reflected as 2.5 percent in the Spring Forecast. 9:10:42 AM Mr. Painter considered slide 6, "LFD Modeling Baseline," which had two graphs showing a fiscal model of the baseline. He pointed out the surplus/deficit listed on the top of the slide and pointed out a surplus of $279 million starting in FY23, a statutory PFD payout of $2.76 billion, with a deficit of $1.5 billion that would decrease overtime to $1 billion by FY27. He relayed that the blue bars on the UGF Budget/Revenue graph denoted traditional revenue, the green bars was the POMV from the Permanent Fund, the yellow bars reflected draws from the Constitutional Budget Reserve (CBR) and Statutory Budget Reserve (SBR), the red bars were unplanned draws form the Earnings Reserve Account (ERA) necessary to balance the budget. The solid line showed the budget in the scenario, the dotted line we the budget minus the dividend. Co-Chair Stedman asked about American Rescue Plan Act (ARPA) funds. Mr. Painter explained that the federal government had provided an unusually flexible source of funds, due to the pandemic. The ARPA had given the state $1 billion to spend on items related to pandemic preparedness, economic response to the pandemic, as well as replacing lost revenue due to the pandemic. In FY 22, the legislature had used $250 million of the funds to replace lost revenues, of which there was a small surplus at the end of FY22. He furthered that the governor had proposed to use another $375 million for that purpose in FY23. Co-Chair Stedman requested that Mr. Painter refrain from using acronyms. 9:13:04 AM Mr. Painter addressed the graph on the right, which showed budget reserves and savings accounts. The yellow bars at the bottom reflected the CBR and SBR. He noted that the assumption was that $500 million, minimum, would be left in the CBR to allow for cashflow for the state. He said that the default assumption was that the CBR would not fall below $500 million, rather the ERA would be used to fill the deficit. He relayed that the green bars showed the spendable amounts of the ERA, which was not a true budget reserve, but was available for the POMV draw or to fill deficits. He shared that along the bottom was the effective POMV draw rate, which was meant to put the size of the overdraw into perspective with the overall size of the draw; the black numbers reflected the 5 percent draw, and the black numbers indicated the overdraw. 9:14:44 AM Mr. Painter displayed slide 7, "Senate Finance Committee Scenarios": Senate Finance Co-Chairs asked for modeling with the following assumptions that differ from LFD's baseline: ?Capital budget baseline of $250 million (instead of $154.7 million) ?Agency operations growing at 2.5% (instead of 2.0%) ?Assume expiring federal funds are replaced with UGF and PERS healthcare is funded after FY23 ?Varying PFD scenarios: statutory PFD, 50/50 of POMV, 75/25 of POMV Mr. Painter noted that the scenario with a capital budget baseline of $250 million incorporated more deferred maintenance and matching funds for the federal infrastructure bill. He noted that the governor had proposed funding the Alaska Marine Highway System with federal funds, which would run out in 5 years and would have to be replaced general funds. He spoke to the PFD scenarios. He noted that the 75/25 of POMV scenario was similar to the version of SB 26 that the committee passed in 2017. 9:16:52 AM Senator Wilson asked whether the capital budget would decrease after the federal match funds expired in 5 years. Mr. Painter answered in the negative. He elaborated that some of the matching funds would be temporary. He said that the majority of the increase was presumably for deferred maintenance. 9:17:50 AM Mr. Painter highlighted slide 8, " Comparison of Senate Finance Committee Scenario to LFD Baseline," which showed a table listing dollar amounts and a bar graph. He shared that the numbers reflected did not include the dividend payout. He related that in the first year the difference was the larger Capital Budget assumption going from $154.7 million to $250 million, in the following years the numbers were a combination of the larger Capital Budget, UGF for expiring federal funds, and the higher growth rate. He said the jump between FY26 and FY27 was when the federal AMHS funds would run out. He said that the assumption was that the funds would have to be replaced with UGF in FY27. He explained that the increasing gap reflected on the slide was due to the half a percent increase in agency operations. He noted that .5 percent did not seem like a lot but ended up being a substantial difference over a 10- year timespan. Co-Chair Stedman understood the 2 percent was what was used by LFD in the scenario. Mr. Painter answered "yes." He said that the division used 2 percent, the slide reflected 2.5 percent, and the governor used 1.5 percent. 9:19:45 AM Mr. Painter looked at slide 9, "Stress Tests": ?Two types of stress tests performed: Budget stress test: grow agency operations and capital budget by 3.5% per year instead of 2.5% Revenue stress test: use probabilistic modeling to simulate a range of possible oil prices and investment returns ?For each PFD scenario, we will show the non- stressed model output and the two stress tests Co-Chair Stedman presumed that the non-stress test was a linear extrapolation without variability in the economy or financial markets. Mr. Painter answered affirmatively. He noted that the division did not have to be a combined stress test; there was no budget and revenue stress test. He said that in the real world it was expected that the legislature would respond to changes in revenue, spending less when revenue was low and more when revenue was high, that budget lines would stay constant so that the impact of policy could be seen as revenue varied. 9:21:40 AM th Mr. Painter thought that slide 10 was an example of a 25 percentile scenario. CONOR BELL, FISCAL ANALYST, LEGISLATIVE FINANCE DIVISION, th Mr. Bell addressed slide 10, " Stress Test: 25 Percentile Example": ?Example of a single case, for which 25% of total cases see greater overall deficits. ?Example case has average oil price of $58 and average Permanent Fund Mr. Bell stated that 2000 different trials were run for each scenario, using different assumptions for oil prices, oil production, ad permanent fund market returns. He stated that the intent was to get a sense of the range of different possibilities. He relayed that there was a lot of sensitivity as to how the inputs were set; oil prices were set by historical variations over time. He related that DOR based their out years on options markets. He stated that the slide showed a single trial, out of 2000, and based on the metric of total deficits run, 25 percent of trials looked worse, and 25 percent looked better. The chart showed the left reflected the oil price over time, with the price of oil averaging out to approximately $58/bbl. He said the other chart showed the Permanent Fund total return. He relayed that FY23 showed a negative 7 percent return, with several lackluster years before regaining strength in the later years. He said that the average return of 5.4 percent was not much lower than Callans projected return of 6.4 percent. He said that the poor performance in the early years meant that there was no compounding growth, which was not ideal. 9:25:49 AM Senator von Imhof appreciated the clarification that 25 percent of the total cases were worse than shown on the slide, and 75 percent of cases were better. 9:26:43 AM Mr. Bell advanced to slide 11, " Scenario 1: Statutory PFD Normal Model Output," which showed two graphs from the output described by one of the scenarios using the SFC baseline of 2.5 percent agency operations inflation assumption and the larger $250 million Capital Budget starting FY23. The first several years showed about $1.5 billion deficits. He pointed out the red bars on the left indicating ERA overdraws. He thought in practice the legislature could react in the situation by cutting budgets or raising revenue, but for modeling purposes it was assumed the gap would be filled with the ERA. Mr. Bell pointed out that that the ERA dropped to $6 billion over time, and one reason for the resiliency of the ERA after a yearly deficit was the strong unrealized gains balance of the Permanent Fund that supplemented the account. He explained that the effective draw rates were 7 percent under the scenario. 9:29:21 AM AT EASE 9:29:45 AM RECONVENED Senator Hoffman noted that in FY23, there was substantial draw from the CBR which would require a three-quarters vote of the legislature. He thought the public should be aware that the ERA could be accessed by the legislature with a simple majority. Co-Chair Stedman asked Mr. Bell to explain the small amount left in the CBR reflected on the slide Mr. Painter explained that currently there was a bit over $1 billion in the CBR. He said that the scenario assumed a surplus in FY22 of $279 million before money was transferred from revenue replacement, the American Rescue Plan, and the SBR. He explained that the scenario assumed there would be no supplemental dividend spending in FY22, so the surplus repaid the CBR for spending in FY23. Co-Chair Stedman presumed that the intent was to maintain a half billion in the CBR, as was recommended by OMB. Mr. Painter answered affirmatively. 9:32:04 AM Senator von Imhof discussed unrealized revenue versus realized revenue. She thought unrealized gains calculated today could possibly not be there in the years to come. She commented on the success in the recent market. She asked whether the fiscal gap could close in a few years. Mr. Painter replied in the affirmative. He shared that the state had two kinds of unrealized gains: real estate that was illiquid, which would be difficult to realize in full value; equities could be sold without impacting value and private equities, which were often long-term investments. There was a mix of investments that yielded various returns on various timelines. He said that it would be difficult to value assets until they were realized. 9:34:03 AM Senator von Imhof did not feel comfortable assuming that the decline reflected by the green bar would be boosted by the sale of state assets. She asserted that the unrealized potential that existed today might not be there in 2027. Mr. Painter thought Senator von Imhof made a good example of why LFD had changed to the probabilistic modeling for the stress tests. He thought switching to that modeling allowed for the capture of the volatility in the Permanent Fund returns in the unique situation where there was such a large unrealized balanced. He stated that the run up in the market of the pervious year had led to an unprecedented unrealized balance. He shared that the probabilistic modeling was more realistic than using historical numbers. 9:35:39 AM Mr. Bell looked at slide 12, "Scenario 1: Statutory PFD Budget Stress Test," which showed two graphs with a statutory PFD and used the SFC baseline assumptions. He noted that the agency operations were reflected as growing at 3.5 percent, which created a larger delta in the later years. He observed that the ERA was drawn down significantly to below $5 billion in FY31. Co-Chair Stedman commented that there was currently an inflation spike of 7.5 percent. He thought there was clearly upward pressure on inflation, and the 1 percent had been added to get an idea of the variability. 9:38:27 AM Mr. Bell showed slide 13, "Scenario 1: Statutory PFD Revenue Stress Test," which was the probabilistic modelling running 2,000 different scenarios, selecting oil price for each year and Permanent Fund returns randomly from a bell curve distribution. The top of the chart showed the median surplus deficit, which was slightly different than the scenario on slide 11. He discussed the variation in numbers at high oil prices versus low oil prices. He commented on the chart on the left, which showed the range of fiscal year end realized ERA balances in millions. The blue line was the same as the surplus/deficit numbers th on the top of the slide, and the yellow bars were the 25 th and 75 percentile. He relayed that 50 percent of the cases were within and 50 percent were outside of the yellow th lines. He said that the vertical back line showed the 10 th to 90 percentile. He noted that the distributions were sensitive to the inputs. He shared that DOR was comparing options to future markets, which gave them a wider price span. He reminded the committee that the slide was an illustration of potential outcomes. 9:42:43 AM Mr. Bell addressed the chart on the right of slide 13, which showed the range of realized ERA balances. He discussed the range of percentiles and noted that the ERA would be drained by FY28. On the high end there was a possibility for the ERA balance to remain at $20 billion; the graph showed the range of uncertainty when trying to predict what would happen in the future. Mr. Bell drew attention to the bottom of the slide, which had a table showing CBR balance possibilities. He said that the first row showed the likelihood of the CBR balance falling below $2.5 billion. The bottom line depicted the probability of the recommended $500 million in the CBR balance over time. He stated that approximately 75 percent of the time in the first two years, the CBR went below the minimum amount. He stated that the gap would need to be filled with some other source once the CBR hit $500 million. 9:45:42 AM Co-Chair Stedman asked for a definition of the median. Mr. Bell explained that the if there were 100 different instances of something, the median was whatever the value is of number 50. The median can be different than the mean if the distribution was skewed. He said that in many of the scenarios the mean and median value differed. 9:46:43 AM Mr. Bell referenced slide 14, "Scenario 2: 50/50 PFD Normal Model Output," which included 2.5 percent agency operations growth. The deficits were significantly smaller because of the smaller PFD amount. The ERA was overdrawn beginning in FY25 but buoyed by unrealized gains. He mentioned that LFD assumed a constant function for the realizing of earnings which was based on 2 percent of the prior years total fund balance, 20 percent of the prior years unrealized gain balance, and 20 percent of a given years total returns. The function created somewhat more certainty. 9:48:42 AM Mr. Bell turned to slide 15, " Scenario 2: 50/50 PFD Budget Stress Test," which showed graphs depicting the growth of the surplus/deficit in millions, the UGF budget/revenue in millions, the budget reserves fiscal year ending balance in millions, and the effective POMV draw rate through FY31. He pointed out that the ERA was overdrawn beginning in FY 25, and there was a greater draw down of the ERA. 9:49:29 AM Mr. Bell considered slide 16, "Scenario 2: 50/50 PFD Revenue Stress Test," which showed what the governor was proposing in his current budget. He said that it was entirely possible that there was a surplus under the model, and it was also possible the there would be fiscal gaps that would need to be filled. He pointed to the realized ERA balance which reflected a 10 percent likelihood of the ERA going to zero, which meant there would be no money from the Permanent Fund that could be spent constitutionally. He furthered that the POMV would not be drawn for government th services. He said that under 25 percentile the ERA emptied out in FY31, meaning there would be a 1 in 4 chance that the ERA would de drawn down entirely by that point. He discussed inflation proofing the said that LFD assumed that it only occurred when there was enough money left in the ERA to pay the current and following years POM draw. 9:52:54 AM Co-Chair Stedman thought it was important to remember that if the ERA went to zero there would be no PFD. He had requested the chart on the right, which he thought was a good way to get a high-level feel of the risk level. 9:54:10 AM Senator von Imhof appreciated the bugle charts Co-Chair Stedman had requested be included on the stress test slides. She thought that the 2000 trials provided a more accurate forecast. She observed that the median was the most common number, which she noted held steady or declined on the slides. She considered the impact of other consequences and further policy the legislature would be forced to make to backfill deficits, which could compound the economic impact. 9:55:48 AM Mr. Bell displayed slide 17, "Scenario 3: 75/25 PFD Normal Model Output," which showed the normal output with the 2.5 percent agency operations growth with the POMV at 25 percent. There were small to moderate surpluses in the first several years. There were no overdraws of the ERA during the period. He noted that right hand chart showed substantial growth in the CBR/SBR and realized ERA. Under the assumptions there was a significant budget reserve nest egg, with substantial growth in the ERA. Co-Chair Stedman explained that the statutory dividend was roughly $4,200 per person. He furthered that a 50/50 dividend was roughly $2,600. He continued that under the 75/25 split depicted on the slide, the dividend would be $1,300. He thought that it was important for the public to realize the scale of individual dividends. 9:58:38 AM Co-Chair Bishop complimented Co-Chair Stedman for establishing for the public the definition of median. He considered that the slide reflected for the first time the ERA meeting the guideline of four and a half times the draw amount in reserves. 9:59:31 AM Senator von Imhof thought it was refreshing to see that the 75/25 scenario working on slide 17 without deficits. She thought it was important to include the total cost to the state treasury when discussing the amount of the PFD for each Alaskan resident. She asked about the yellow bars showing the CBR/SBR on the right-hand graph and pondered whether it presumed a spending cap and excess money was being put in the CBR rather than agency funding. She asked if the model assumed the reverse sweep did not occur and all those funds went into the CBR. Mr. Painter relayed that the scenario assumed the baseline budget with agency growth at 2.5 percent. The scenario assumed there was no reverse sweep, as it was not considered in the governors budget. He said that regardless of the reverse sweep, the surpluses went back to the CBR by default. Mr. Painter continued that when the legislature had surpluses in the past, they did not always direct it to the CBR. He concluded that in the case of the model on the slide the assumption was, absent of any policy call by the legislature, the money would be deposited into the CBR. 10:02:16 AM Co-Chair Stedman interjected that when the remaining account balances were transferred, or swept, at the end of the year, any money owed to the CBR was repaid. He said that the sweep was to get the funds back out and use the funds for the beginning of the fiscal year balanced for those sweepable funds. 10:02:56 AM Mr. Bell highlighted slide 18, "Scenario 3: 75/25 PFD Budget Stress Test," which showed the same 75/25 PFD scenario. He noted that there were still surpluses, but eventually they went to zero in the out years. He pointed out the absence of ERA overdraws and the limited CBR growth. He stated that the ERA was unaffected and behaved in the same way as the scenario on slide 16. 10:04:11 AM Mr. Bell looked at slide 19, "Scenario 3: 75/25 PFD Revenue Stress Test," which showed $200 million in surpluses at the median level, approximately $500 million in deficits in the early years at the 25 percent level, and nearly $1.5 billion surplus at 75 percent. He noted the wide band of uncertainty reflected on the slide. He stated that the expected surplus did rise in the later years. He noted that the ERA could go to zero even without overdraws, simply based upon poor performance of the Permanent Fund. He offered that there was about a ten percent chance that the ERA was entirely depleted by FY29. 10:06:05 AM Senator von Imhof wanted to marry the two charts. She thought the chart on the left showed the median was above zero from FY23 to FY 31. She understood that the median was then reflected on the thick black line on the bugle chart on the right-hand side. She pondered that whether the state was running at a surplus in the median then the ERA would remain stable. Mr. Bell thought the graph showed the risk of volatility. The previous year LFD had shown volatility of the returns by using historical returns and projecting them upon future scenarios. He noted that the median line represented the median amount for every year. He continued to discuss the possible scenarios and the varying fiscal outcomes of the projected numbers. 10:08:55 AM Co-Chair Bishop asked about the inflation rate as shown on slide 18. Mr. Bell stated that agency operations growth in the budget stress tests was either 2.5 or 3 percent. He said that the model assumed inflation for the principal inflation proofing at 2 percent. He said that the assumption was that the budget was growing faster than the inflation proofing transfer. Co-Chair Bishop asked for the inflation rate for scenario 3, at the normal model output. Mr. Bell stated that the model used 2.5 percent for agency operation growth and 2 percent for inflation. Co-Chair Bishop asked whether the stress test on slide 18 was running at 3.5 percent. 10:10:31 AM Senator Hoffman looked at slide 18 and observed that the CBR was still owed $18 billion from the state. Mr. Painter stated that the number was subject to dispute between executive branch and the Division of Legislative Audit. Senator Hoffman explained that the action had been taken in the past to substantially pay back the CBR and bolster state savings. 10:11:52 AM Mr. Painter commented on Senator von Imhof's earlier question. He stated that one of the reasons that the median had the ERA shrinking slightly was that the 6.2 percent return and the 2 percent inflation resulted in only a 4.2 percent return above inflation. He said that a 5 percent draw of the fund would drain the fund. Co-Chair Stedman said that other scenarios could be considered. Co-Chair Stedman thanked Mr. Painter and Mr. Bell for the time spent on modelling the scenarios. He expected that there would be many legislators requesting the slides for study. Co-Chair Stedman discussed housekeeping. ADJOURNMENT 10:14:09 AM The meeting was adjourned at 10:14 a.m.