Moderate inflation, échelle mobile, and all that
In the old days, moderate inflation was inflation that exceeded garden variety, and ordinarily it denoted 20-25 percent annual CPI. The claim is that old school ‘moderate inflations’ either stepped back to garden variety or quickly turned into ‘runaway inflation’. Now, runaway inflation is just another name for hyperinflation but here I use the term to emphasize that 20-30 percent inflation is neither high and chronic nor does it comfortably provide a persistent plateau. It will either fall to single-digits or jump to a higher plateau, and if this happens, voilà! Back to the 1990s, in a sense. Could this happen? As domestic producer prices are now a clear and present danger, and the risk of a 55-60 percent D-PPI is very real, could then CPI also rise to a much higher plateau?
D spers on and pr c ng behav our
Annual CPI now stands at 24.52 percent whereas annual D-PPI at 46.22 percent. There is a large gap. Producer prices reflect costs, and therefore they go up fast as the exchange rate depreciates. What we call the exchange rate pass-through effect isn’t only the direct impact of the exchange rate, it also includes imported intermediate goods price changes. Import prices, including energy prices, add to the direct impact. Now, according to the Diebold-Yilmaz Diffusion Index, the rate at which producer inflation translates to consumer inflation rose. The rather fast-moving multiple at which such a translation occurs is dependent on both oil prices and the exchange rate. As inflation rises, so does the diffusion parameter. I can add to this a ‘residual’, which one might call inflation memory and/or pricing behaviour parameter. To the extent economic actors believe that they face a prolonged period of high inflation or an extended time frame during which the lira might not stabilize, they will try to pass on production cost increases to consumer prices faster and at a higher rate. Although the approximate 46 percent PPI versus 25 percent CPI gap won’t automatically close, meaning CPI rises very steeply to catch up with the PPI, this time around there is “accumulated inflation” that awaits us in 2019, so to speak. Even if the lira stabilizes, as it stands, the already incurred production cost burden will show in the CPI in the coming months. 30-35 percent CPI is now in the cards regardless of lira stabilization.
There are two factors that might help curtail CPI hikes. First, domestic demand could so swiftly collapse that producers would be unable to reflect cost pressures on retail prices, in which case they will produce for turnover alone. Not even EBITDAs will be secure. Second, the lira could not only stabilize but it could actually gain value, signaling such a clear and credible new programmatic approach that market players believe there is a sea-change. Otherwise, I would expect to see over 30 percent CPI by early 2019.
And not only the d ffus on ndex…
Zafer Yükseler’s Fast Consumed Products Index (FCPI) has posted 16.48 percent increase yearto-date, and 21.49 percent annually. It is a sub-index, an index of products that are consumed at a higher frequency and collects 138 items out of 407 that add up to the CPI itself. The most significant news on that score is the monthly September FCPI changes: 0.99 percent in September 2017 versus 5.13 percent in September 2018. This is indeed a huge difference and shows how fast inflation has been spiralling up. It will no doubt rise further in the coming months. Seasonally-adjusted service inflation rose from 9.24 percent in January to 13.94 percent in September. Now the C-core inflation – excluding energy, food and non-alcoholic beverages, alcoholic beverages, tobacco and gold - has reached 23.82 percent, up from 12.21 percent in January. It has effectively doubled within 8 months. Food prices are a disaster, up from 8.76 percent in January to 27.68 percent in September. Now these are all seasonally-adjusted indices. The graphics depict the co-movements of the exchange rate, D-PPI and CPI, core inflation and CPI. They all move up with an almost perfect correlation and they are all tied up with the exchange rate. Note that natural gas, electricity and pump prices are all awaiting their turn. Contributions show a widespread movement, with food, housing, houseware and transportation prices rising most swiftly. Housing construction cost increases are expected to reach 30 percent. Diffusion, level, cross-correlations across sub-items all point to a composite phenomenon. Should oil prices climb to $100 a barrel there will be additional pressures, and if the lira is hit again, there will be a renewed inflationary ac-
cumulation deferred to the coming months. The pass-through from the exchange rate is about 15 percent on average, which means a 6 percentage point addendum to the already existing inflation trend. As such, if there were the exchange rate alone, we could expect 19-20 percent year-end inflation, but it isn’t only that, as mentioned above. As inflation expectations deteriorate and as there is no clear indication to the effect that the lira will stabilize for good, pricing behavior begins to reflect stickiness, not only downwards but also as a persistent forward-looking pattern. This I may call an ‘automatic forward-looking inflationary self-spiral’. Although widespread, the origin lies in the manufacturing sector’s cost increases. Automobiles, TVs, white goods and all other consumer durables’ prices are rising quasi-automatically by 15-20 percent, and over 10 percent monthon-month. This is indeed worrisome. The rise in the processed food sub-item that has reached 6.4 percent in a single month is also attributable to the exchange rate. I repeat: barring any further exchange rate shock, the average CPI inflation that lies dormant 6-9 months ahead is about 28 percent, with a peak that is most probably above 30 percent. Now, it is way above my June expectation of 22 percent. This is what I mean by ‘runaway inflation’, not hyperinflation but an “inflation stock” that piles up continually. No wonder TRY deposit rates are now around 25 percent, and lending rates around 35 percent. Even this may not provide a buffer against inflation because we are talking about 28 percent average CPI that may lie ahead.
Inflat on accommodat on, memory format on and échelle mob le
The question is: what to do with all this? If you accommodate in- flation, what you have to do is simple. Print money if need be, and raise wages accordingly. Now, 2535 percent wage increases aren’t something Turkish businesses can handle at this point. The state can do that, but if it does it will accommodate inflation. If you don’t do that, obviously real wages and salaries will erode fast, and people will feel that on the spot. If you do that, you will index inflation. Now, backward indexation creates or at least adds momentum to inflation memory, and renders it persistent. An extreme form of this is the infamous échelle mobile. The tricky thing here is to determine whether inflation is here to stay or is it a one-off event. Obviously, it isn’t a one-off event, which leaves little choice. The choice is the optimal level of wage adjustments. Perhaps forward-indexation is a good idea. This way, as credibility builds up anew and inflation expectations fall, wage adjustment multiples also fall. Anyway, wages are in general procyclical in this country and inflation never builds up because of wage increases. Except the pivotal minimum wage that creates a zone of a-cyclicality across the business cycle, wages haven’t been a cause for concern over almost three decades.
This is no Argentina. No, there hasn’t been an excess demand that kept inflation always high either. Inflation is in its very heart is a cost phenomenon; therefore, an exchange rate phenomenon, for two reasons: Imported intermediate goods matter and FX debt matters. Material balances’ costs and financing costs drive producer prices. Consumer prices follow not because there is always excess demand at the margin – lack of market-clearing at infinity, which is a theoretical possibility - but to the extent demand is strong enough to allow producers to reflect costs on to retail prices. The rest requires a symptom-laden reading - what inflation in reality stands for, its relationship with the current account deficit and suchlike, a long story.
Hence, disinflation means a lot for Turkey because inflation has deeper roots than an ordinary demand-supply outlook would suggest. In the late 1980s and 1990s, the Turkish economy had developed long inflation memory that persisted until the recent past to some extent, but it surely persisted until 2005. Disinflation was key then, and it is key today also. However, in the decade from 2005 to 2015, inflation didn’t really fall as it should have. True, from 8 percent to 5 percent there was a long way. Nobody attempted this because it would have required a genuine attempt that warranted an engineered intervention to the productive structure of industry and to the cost structure of intermediate imports and what not. It would have been politically costly also. Hence, so long as inflation hovered around a 7 percent annual trend, nobody perceived it as a priority. Stabilizing inflation somewhere below 5 percent wasn’t an easy task anyway. Even if monetary and fiscal policies functioned well in tandem, such a process could be still reversible, given adverse exogenous factors, and inflation might have picked up anew and reached the quicksand of double digits. But there was no attempt at reducing inflation permanently below 5 percent while there was plenty of time and opportunity. Half full was the glass. Now the halfbaked disinflation of 2004-2005 has come back with a vengeance. “Why now?” deserves an answer that points not only to external factors - politics, the exchange rate, energy prices, corporate sector FX debt etc. - but also to purely domestic factors, and to the bearers of a long-recognized, time-honoured tradition of the now yet again (in)famous cognitive deficiency syndrome. This syndrome can quickly become a contagious phenomenon in this country.