Gimar special topic edition the impact of climate change on the financial stability of the insurance sector

Figure 2:
NGFS scenarios
Source: NGFS (2021)
Tr
ansition risks
Lo
w
High 
Disorderly
Orderly
Low 
High
Too little, too late
Hot house world
Physical risks
Divergent 
Net Zero 
(1.5°C)
Net Zero
2050 
(1.5°C)
Below
2°C
Delayed 
transition
NDCs
Current 
policies

---

30
The scenario analysis in this report estimates the 
value of insurers’ current asset portfolios under 
an orderly transition, a disorderly transition and a 
“too little, too late” world. As the report focuses 
on transition risks, there is no separate analysis 
of a “hot house” or purely physical risk scenario. 
Instead, physical risk factors consistent with a hot 
house scenario are embedded in the “too little, too 
late” scenario.
4.1.2 Stress factors design for equity,
corporate bonds, and loans and mortgages
As noted earlier, the data assessed for this report 
indicates that about 35% of insurers’ assets on 
average are held in equities (including in investment 
funds), corporate bonds, and loans and mortgages.
The effects of these adverse scenarios on the 
market value of insurers’ assets are assumed to 
exhibit sectoral heterogeneity. For sector-level 
analysis, this report relies on scenario-consistent 
stress factors derived and applied in both 
the academic sphere and among supervisory 
authorities.
39
The stress factors consist of an 
arithmetic mean for each asset class and each 
climate-relevant sector, from those used in publicly 
available methodologies.
4.1.2.1 Orderly and disorderly transition 
scenarios
Equities
The range of the stress factors varies with respect 
to the asset class and sectoral segmentation. 
Graph 11 shows that most sectors exhibit 
moderate variance across the methodologies, 
although utilities shows greater uncertainty. 
By using information from a wide range of 
methodologies, less weight is given to any single 
stress factor — some of which may contain 
intentionally severe assumptions that are difficult 
to compare with those made in other studies.
Cristal, if possible can we stretch the legend from left-hand graph across both graphs (or remove and paste below)? It applies to both.
0
100
200
300
400
500
600
700
800
900
2015
2020
2025
2030
2035
2040
2045
2050